physicalean- research 2022

Reservoir of Carbapenemase-Producing Organisms ® Downloaded from mbio.asm.org on February 10, 2018 - Published by mbio.asm.org
mbio.asm.org 13
transmission to patients, as few clonal isolates were detected in patients and the

environment. One might expect to see more transmission if there were a breakdown of

good practices. The diversity of CPO-associated plasmids and transposon elements

from the environment and the isolation of similar plasmid backbones with and without

the blaKPC gene prompt a greater appreciation of the extensive nature and mobility of
plasmids in the environment.

MATERIALS AND METHODS

Setting and sample collection. Environmental samples were collected between January 2012 and
December 2016 at the NIHCC (Bethesda, MD), a 200-bed clinical research hospital (Table 1). Seven
wastewater samples were collected from two manholes outside the hospital, and nine wastewater
samples were collected from the pipes in interstitial spaces serving the ICU and non-ICU rooms. Water
samples were collected from faucets and traps of sinks in patient rooms and common areas. Surfaces
were sampled with BBL culture swabs (Becton, Dickinson and Company, Franklin Lakes, NJ) moistened
with saline or gauze incubated in 10 ml of tryptic soy broth (TSB; Hardy Diagnostics, Santa Maria, CA).
Approximately six inches of pipe immediately below the surfaces of the sink drains was sampled with
moistened BBL culture swabs (Becton, Dickinson and Company).

Sample processing, identification, and characterization of bacterial isolates. Swabs were trans-
ferred into 1 ml of TSB (Remel Inc., Lenexa, KS), vortexed, and plated directly. Gauze in 10 ml of TSB was
incubated overnight at 35°C in 5% CO2 prior to culture. Wastewater samples were plated directly. One
drop of each sample suspension was cultured on CRE, ESBL, and R2A (Remel Inc.) media. A subset of
environmental samples was also cultured on 5% sheep blood agar (SB; Remel Inc.). Perirectal surveillance
cultures were plated directly onto CRE medium. Clinical cultures were plated onto routine clinical
microbiology media, which included a combination of SB, chocolate agar (Remel Inc.), MacConkey agar
(Remel Inc.), and/or Fastidious Broth (Hardy Diagnostics; Fig. S1). HardyCHROM media were incubated
overnight at 35°C in ambient air, and R2A medium was incubated at 30°C in ambient air. Clinical cultures
were incubated at 35°C in 5% CO2. Colonies with different morphologies were isolated and screened by
PCR for the presence of blaKPC or blaNDM (http://www.cdc.gov/HAI/pdfs/labSettings/KPC-NDM-protocol
-2011.pdf). All isolates that tested positive for blaKPC or blaNDM by PCR were identified by MALDI-TOF MS
(BioTyper; Bruker, Billerica, MA). Susceptibility testing was performed via broth microdilution (Trek
Diagnostics, Oakwood Village, OH, or BD Phoenix, Franklin Lakes, NJ, depending on the date of isolation)
or disk diffusion (Thermo Fisher Scientific, Lenexa, KS). Susceptibility interpretations were based on the
most recent Clinical and Laboratory Standards Institute (CLSI) breakpoints (62) (Table S1). For this study,
given our environmental screening approach, we classified as a CPO any isolate that was blaKPCϩ or
blaNDMϩ by PCR. For the antimicrobial susceptibility comparison of members of the family Enterobacte-
riaceae and to eliminate bias potentially caused by the use of different culture media for clinical and
environmental testing, blaKPCϩ or blaNDMϩ environmental Enterobacteriaceae isolates from ESBL or R2A
medium that were susceptible to two or more carbapenems were tested for growth selectivity on CRE
medium in accordance with CLSI M22-A3 guidelines (63). Briefly, a 0.5 McFarland suspension in saline was
prepared and diluted 1:10. Ten microliters was inoculated onto CRE medium and incubated at 35°C in
ambient air for 18 to 24 h. Only isolates that grew on this confirmatory CRE agar screen were included
in the susceptibility comparison.

Genome sequencing and assembly. Genomic DNA extraction and library preparation for Illumina
MiSeq and single-molecule real-time (SMRT) sequencing were performed as described previously (64, 65).
Paired-end MiSeq reads were assembled into contigs by using SPAdes version 3.5.1 (66) and further
polished by using Pilon (67). Scaffolding of refined genome assemblies to reference genomes was
performed by using ABACAS version 1.3.1 (68). PacBio RSII SMRT sequencing reads (P6 polymerase
binding and C4 sequencing kit, 240-min collection) were assembled by using HGAP3 (PacBio SMRT
Analysis version 2.3) or CANU (version 1.3, 1.4, or 1.5) (69). The resulting contigs were polished by using
Quiver (SMRT Analysis version 2.3) and manually reviewed and finished by using Gepard (version 1.3)
(70). Base calling and finishing accuracy was assessed by using the Most Probable Genotype (71) software
with Illumina MiSeq read alignments against the finished PacBio assembly. The Integrative Genomics
Viewer (72) was used to visualize the MiSeq read alignments. Genome annotation was performed by
using the National Center for Biotechnology Information (NCBI) prokaryotic Genome Annotation Pipeline
(PGAP, https://www.ncbi.nlm.nih.gov/genome/annotation_prok/).

Bioinformatic analyses. Genetic relatedness between genome assemblies was assessed by ANI and
SNPs. ANI was estimated by using Mash version 1.1.1 with a k-mer size of 21 and a sketch size of 1,000
(73), while SNPs were called by using Snippy version 3.2 (BWA-MEM version 0.7.15, freebayes version
1.0.2) (https://github.com/tseemann/snippy). All SNPs required a read depth of Ն10 reads with Ն90% of
those aligned reads supporting the variant nucleotide call. Core genome alignments were performed by
using Parsnp and visualized by using Gingr (48).

Plasmid detection in shotgun assemblies was determined by using a k-mer inclusion approach.
Briefly, 16-bp k-mers were generated from the fully resolved PacBio plasmid sequences along with the
MiSeq genome assemblies by using meryl (13 May 2015 release, http://kmer.sourceforge.net/). A plasmid
was considered present within a genome assembly if Ն95% of the plasmid k-mers were contained
within the MiSeq genome assembly k-mers. A similar k-mer inclusion analysis of the PacBio-generated
blaKPC-positive and blaKPC-negative plasmid sequences was performed. Plasmid coverage and align-
ments were computed by using NUCmer version 3.1 (74), laid out by using a custom Perl script, and
visualized in R.

January/February 2018 Volume 9 Issue 1 e02011-17

Weingarten et al. ® Downloaded from mbio.asm.org on February 10, 2018 - Published by mbio.asm.org

Data and material availability. This report references genomic data from previous studies
(PRJNA251756). Genomic data generated for this report have been deposited under project no.
PRJNA430442. Isolates can be obtained from K.M.F.; a material transfer agreement is necessary.

SUPPLEMENTAL MATERIAL
Supplemental material for this article may be found at https://doi.org/10.1128/mBio

.02011-17.
FIG S1, PDF file, 0.01 MB.
FIG S2, PDF file, 0.02 MB.
FIG S3, PDF file, 0.05 MB.
TABLE S1, XLSX file, 0.1 MB.
TABLE S2, XLSX file, 0.04 MB.

ACKNOWLEDGMENTS
We thank the medical technologists of the NIH Clinical Center Department of

Laboratory Medicine Microbiology Service for their many contributions described in
this report. We thank J. Fekecs and D. Leja for providing graphic design assistance with
figures. This study utilized the high-performance computational capabilities of the NIH
Biowulf Linux cluster.

This study was supported by the National Human Genome Research Institute,
National Institute of Allergy and Infectious Diseases, and the NIH Clinical Center
Intramural Research programs and by an NIH Director’s Challenge Innovation Award.

REFERENCES 15(th), 2013. Stand Genomic Sci 8:571–579. https://doi.org/10.4056/sigs
.4187859.
1. Magill SS, Edwards JR, Bamberg W, Beldavs ZG, Dumyati G, Kainer MA, 8. Wilson AP, Smyth D, Moore G, Singleton J, Jackson R, Gant V, Jeanes A,
Lynfield R, Maloney M, McAllister-Hollod L, Nadle J, Ray SM, Thompson Shaw S, James E, Cooper B, Kafatos G, Cookson B, Singer M, Bellingan G.
DL, Wilson LE, Fridkin SK, Emerging Infections Program Healthcare- 2011. The impact of enhanced cleaning within the intensive care unit on
Associated Infections and Antimicrobial Use Prevalence Survey Team. contamination of the near-patient environment with hospital
2014. Multistate point-prevalence survey of health care-associated in- pathogens: a randomized crossover study in critical care units in two
fections. N Engl J Med 370:1198 –1208. https://doi.org/10.1056/NEJM hospitals. Crit Care Med 39:651– 658. https://doi.org/10.1097/CCM.0b013
oa1306801. e318206bc66.
9. Kotsanas D, Wijesooriya WR, Korman TM, Gillespie EE, Wright L, Snook K,
2. Tamma PD, Goodman KE, Harris AD, Tekle T, Roberts A, Taiwo A, Simner Williams N, Bell JM, Li HY, Stuart RL. 2013. ‘Down the drain’: carbapenem-
PJ. 2017. Comparing the outcomes of patients with carbapenemase- resistant bacteria in intensive care unit patients and handwashing sinks.
producing and non-carbapenemase-producing carbapenem-resistant Med J Aust 198:267–269. https://doi.org/10.5694/mja12.11757.
Enterobacteriaceae bacteremia. Clin Infect Dis 64:257–264. https://doi 10. Leitner E, Zarfel G, Luxner J, Herzog K, Pekard-Amenitsch S, Hoenigl M,
.org/10.1093/cid/ciw741. Valentin T, Feierl G, Grisold AJ, Högenauer C, Sill H, Krause R, Zollner-
Schwetz I. 2015. Contaminated handwashing sinks as the source of a
3. Karlowsky JA, Lob SH, Kazmierczak KM, Badal RE, Young K, Motyl MR, clonal outbreak of KPC-2-producing Klebsiella oxytoca on a hematology
Sahm DF. 2017. In vitro activity of imipenem against carbapenemase- ward. Antimicrob Agents Chemother 59:714 –716. https://doi.org/10
positive Enterobacteriaceae isolates collected by the SMART global sur- .1128/AAC.04306-14.
veillance program from 2008 to 2014. J Clin Microbiol 55:1638 –1649. 11. Lowe C, Willey B, O’Shaughnessy A, Lee W, Lum M, Pike K, Larocque C,
https://doi.org/10.1128/JCM.02316-16. Dedier H, Dales L, Moore C, McGeer A, Mount Sinai Hospital. 2012.
Outbreak of extended-spectrum beta-lactamase-producing Klebsiella
4. Snitkin ES, Zelazny AM, Thomas PJ, Stock F; NISC Comparative Sequenc- oxytoca infections associated with contaminated handwashing sinks.
ing Program Group, Henderson DK, Palmore TN, Segre JA. 2012. Tracking Emerg Infect Dis 18:1242–1247. https://doi.org/10.3201/eid1808.111268.
a hospital outbreak of carbapenem-resistant Klebsiella pneumoniae with 12. Roux D, Aubier B, Cochard H, Quentin R, van der Mee-Marquet N, HAI
whole-genome sequencing. Sci Transl Med 4:148ra116. https://doi.org/ Prevention Group of the Réseau des Hygiénistes du Centre. 2013. Con-
10.1126/scitranslmed.3004129. taminated sinks in intensive care units: an underestimated source of
extended-spectrum beta-lactamase-producing Enterobacteriaceae in the
5. Conlan S, Thomas PJ, Deming C, Park M, Lau AF, Dekker JP, Snitkin ES, patient environment. J Hosp Infect 85:106 –111. https://doi.org/10.1016/
Clark TA, Luong K, Song Y, Tsai YC, Boitano M, Dayal J, Brooks SY, j.jhin.2013.07.006.
Schmidt B, Young AC, Thomas JW, Bouffard GG, Blakesley RW, NISC 13. Wolf I, Bergervoet PW, Sebens FW, van den Oever HL, Savelkoul PH, van
Comparative Sequencing Program, Mullikin JC, Korlach J, Henderson DK, der Zwet WC. 2014. The sink as a correctable source of extended-
Frank KM, Palmore TN, Segre JA. 2014. Single-molecule sequencing to spectrum beta-lactamase contamination for patients in the intensive
track plasmid diversity of hospital-associated carbapenemase-producing care unit. J Hosp Infect 87:126 –130. https://doi.org/10.1016/j.jhin.2014
Enterobacteriaceae. Sci Transl Med 6:254ra126. https://doi.org/10.1126/ .02.013.
scitranslmed.3009845. 14. Hopman J, Tostmann A, Wertheim H, Bos M, Kolwijck E, Akkermans R,
Sturm P, Voss A, Pickkers P, Vd Hoeven H. 2017. Reduced rate of
6. Clarivet B, Grau D, Jumas-Bilak E, Jean-Pierre H, Pantel A, Parer S, Lotthé intensive care unit acquired Gram-negative bacilli after removal of sinks
A. 2016. Persisting transmission of carbapenemase-producing Klebsiella and introduction of “water-free” patient care. Antimicrob Resist Infect
pneumoniae due to an environmental reservoir in a university hospital, Control 6:59. https://doi.org/10.1186/s13756-017-0213-0.
France, 2012 to 2014. Euro Surveill 21:piiϭ30213. https://doi.org/10 15. Chagas TP, Seki LM, da Silva DM, Asensi MD. 2011. Occurrence of
.2807/1560-7917.ES.2016.21.17.30213.
mbio.asm.org 14
7. Shogan BD, Smith DP, Packman AI, Kelley ST, Landon EM, Bhangar S,
Vora GJ, Jones RM, Keegan K, Stephens B, Ramos T, Kirkup BC, Jr., Levin
H, Rosenthal M, Foxman B, Chang EB, Siegel J, Cobey S, An G, Alverdy JC,
Olsiewski PJ, Martin MO, Marrs R, Hernandez M, Christley S, Morowitz M,
Weber S, Gilbert J. 2013. The Hospital Microbiome Project: meeting
report for the 2nd Hospital Microbiome Project, Chicago, USA, January

January/February 2018 Volume 9 Issue 1 e02011-17

Reservoir of Carbapenemase-Producing Organisms ® Downloaded from mbio.asm.org on February 10, 2018 - Published by mbio.asm.org

KPC-2-producing Klebsiella pneumoniae strains in hospital wastewater. J Lett Appl Microbiol 48:560 –565. https://doi.org/10.1111/j.1472-765X
Hosp Infect 77:281. https://doi.org/10.1016/j.jhin.2010.10.008. .2009.02572.x.
16. Koh TH, Ko K, Jureen R, Deepak RN, Tee NW, Tan TY, Tay MR, Lee VJ, 32. Decker BK, Lau AF, Dekker JP, Spalding CD, Sinaii N, Conlan S, Henderson
Barkham TM. 2015. High counts of carbapenemase-producing Entero- DK, Segre JA, Frank KM, Palmore TN. 2018. Healthcare personnel intes-
bacteriaceae in hospital sewage. Infect Control Hosp Epidemiol 36: tinal colonization with multidrug-resistant organisms. Clin Microbiol
619 – 621. https://doi.org/10.1017/ice.2015.44. Infect 24:82.e1– 82.e4 https://doi.org/10.1016/j.cmi.2017.05.010.
17. Vergara-López S, Domínguez MC, Conejo MC, Pascual Á, Rodríguez-Baño 33. Hughes HY, Conlan SP, Lau AF, Dekker JP, Michelin AV, Youn JH, Hen-
J. 2013. Wastewater drainage system as an occult reservoir in a pro- derson DK, Frank KM, Segre JA, Palmore TN. 2016. Detection and whole-
tracted clonal outbreak due to metallo-␤-lactamase-producing Klebsiella genome sequencing of carbapenemase-producing Aeromonas hydro-
oxytoca. Clin Microbiol Infect 19:E490 –E498. https://doi.org/10.1111/ phila isolates from routine perirectal surveillance culture. J Clin Microbiol
1469-0691.12288. 54:1167–1170. https://doi.org/10.1128/JCM.03229-15.
18. White L, Hopkins KL, Meunier D, Perry CL, Pike R, Wilkinson P, Pickup RW, 34. Muzslay M, Moore G, Alhussaini N, Wilson AP. 2017. ESBL-producing
Cheesbrough J, Woodford N. 2016. Carbapenemase-producing Entero- Gram-negative organisms in the healthcare environment as a source of
bacteriaceae in hospital wastewater: a reservoir that may be unrelated to genetic material for resistance in human infections. J Hosp Infect 95:
clinical isolates. J Hosp Infect 93:145–151. https://doi.org/10.1016/j.jhin 59 – 64. https://doi.org/10.1016/j.jhin.2016.09.009.
.2016.03.007. 35. Yablon BR, Dantes R, Tsai V, Lim R, Moulton-Meissner H, Arduino M,
19. Zhang X, Lü X, Zong Z. 2012. Enterobacteriaceae producing the KPC-2 Jensen B, Patel MT, Vernon MO, Grant-Greene Y, Christiansen D, Conover
carbapenemase from hospital sewage. Diagn Microbiol Infect Dis 73: C, Kallen A, Guh AY. 2017. Outbreak of Pantoea agglomerans blood-
204 –206. https://doi.org/10.1016/j.diagmicrobio.2012.02.007. stream infections at an oncology clinic-Illinois, 2012–2013. Infect Control
20. Ludden C, Reuter S, Judge K, Gouliouris T, Blane B, Coll F, Naydenova Hosp Epidemiol 38:314 –319. https://doi.org/10.1017/ice.2016.265.
P, Hunt M, Tracey A, Hopkins KL, Brown NM, Woodford N, Parkhill J, 36. van Duin D, Doi Y. 2017. The global epidemiology of carbapenemase-
Peacock SJ. 2017. Sharing of carbapenemase-encoding plasmids be- producing Enterobacteriaceae. Virulence 8:460 – 469. https://doi.org/10
tween Enterobacteriaceae in UK sewage uncovered by MinION se- .1080/21505594.2016.1222343.
quencing. Microbial Genomics 3:e000114. https://doi.org/10.1099/ 37. Espinal P, Mosqueda N, Telli M, van der Reijden T, Rolo D, Fernández-
mgen.0.000114. Orth D, Dijkshoorn L, Roca I, Vila J. 2015. Identification of NDM-1 in a
21. Hocquet D, Muller A, Bertrand X. 2016. What happens in hospitals does putatively novel Acinetobacter species (‘NB14’) closely related to Acin-
not stay in hospitals: antibiotic-resistant bacteria in hospital wastewater etobacter pittii. Antimicrob Agents Chemother 59:6657– 6660. https://doi
systems. J Hosp Infect 93:395– 402. https://doi.org/10.1016/j.jhin.2016.01 .org/10.1128/AAC.01455-15.
.010. 38. Hu H, Hu Y, Pan Y, Liang H, Wang H, Wang X, Hao Q, Yang X, Yang X, Xiao
22. Alam MZ, Aqil F, Ahmad I, Ahmad S. 2013. Incidence and transferability X, Luan C, Yang Y, Cui Y, Yang R, Gao GF, Song Y, Zhu B. 2012. Novel plasmid
of antibiotic resistance in the enteric bacteria isolated from hospital and its variant harboring both a bla(NDM-1) gene and type IV secretion
wastewater. Braz J Microbiol 44:799 – 806. https://doi.org/10.1590/S1517 system in clinical isolates of Acinetobacter lwoffii. Antimicrob Agents Che-
-83822013000300021. mother 56:1698–1702. https://doi.org/10.1128/AAC.06199-11.
23. Chandran SP, Diwan V, Tamhankar AJ, Joseph BV, Rosales-Klintz S, 39. Jones LS, Carvalho MJ, Toleman MA, White PL, Connor TR, Mushtaq A,
Mundayoor S, Lundborg CS, Macaden R. 2014. Detection of carbapenem Weeks JL, Kumarasamy KK, Raven KE, Török ME, Peacock SJ, Howe RA,
resistance genes and cephalosporin, and quinolone resistance genes Walsh TR. 2015. Characterization of plasmids in extensively drug-
along with oqxAB gene in Escherichia coli in hospital wastewater: a resistant Acinetobacter strains isolated in India and Pakistan. Antimicrob
matter of concern. J Appl Microbiol 117:984 –995. https://doi.org/10 Agents Chemother 59:923–929. https://doi.org/10.1128/AAC.03242-14.
.1111/jam.12591. 40. McGann P, Milillo M, Clifford RJ, Snesrud E, Stevenson L, Backlund MG,
24. Picão RC, Cardoso JP, Campana EH, Nicoletti AG, Petrolini FV, Assis DM, Viscount HB, Quintero R, Kwak YI, Zapor MJ, Waterman PE, Lesho EP.
Juliano L, Gales AC. 2013. The route of antimicrobial resistance from the 2013. Detection of New Delhi metallo-␤-lactamase (encoded by
hospital effluent to the environment: focus on the occurrence of KPC- blaNDM-1) in Acinetobacter schindleri during routine surveillance. J Clin
producing Aeromonas spp. and Enterobacteriaceae in sewage. Diagn Microbiol 51:1942–1944. https://doi.org/10.1128/JCM.00281-13.
Microbiol Infect Dis 76:80 – 85. https://doi.org/10.1016/j.diagmicrobio 41. Wang X, Zhang Z, Hao Q, Wu J, Xiao J, Jing H. 2014. Complete genome
.2013.02.001. sequence of Acinetobacter baumannii ZW85-1. Genome Announc
25. Betteridge T, Merlino J, Natoli J, Cheong EY, Gottlieb T, Stokes HW. 2013. 2:e01083-13. https://doi.org/10.1128/genomeA.01083-13.
Plasmids and bacterial strains mediating multidrug-resistant hospital- 42. Zou D, Huang Y, Zhao X, Liu W, Dong D, Li H, Wang X, Huang S, Wei X, Yan
acquired infections are coresidents of the hospital environment. Microb X, Yang Z, Tong Y, Huang L, Yuan J. 2015. A novel New Delhi metallo-␤-
Drug Resist 19:104 –109. https://doi.org/10.1089/mdr.2012.0104. lactamase variant, NDM-14, isolated in a Chinese hospital possesses in-
26. Bäumlisberger M, Youssar L, Schilhabel MB, Jonas D. 2015. Influence creased enzymatic activity against carbapenems. Antimicrob Agents Che-
of a non-hospital medical care facility on antimicrobial resistance in mother 59:2450–2453. https://doi.org/10.1128/AAC.05168-14.
wastewater. PLoS One 10:e0122635. https://doi.org/10.1371/journal.pone 43. Hoyos-Mallecot Y, Rojo-Martín MD, Bonnin RA, Creton E, Navarro Marí
.0122635. JM, Naas T. 2017. Draft genome sequence of NDM-1-producing Leclercia
27. Runcharoen C, Moradigaravand D, Blane B, Paksanont S, Thammachote adecarboxylata. Genome Announc 5:e00135-17. https://doi.org/10.1128/
J, Anun S, Parkhill J, Chantratita N, Peacock SJ. 2017. Whole genome genomeA.00135-17.
sequencing reveals high-resolution epidemiological links between clin- 44. Anuradha M. 2014. Leclercia adecarboxylata isolation: case reports and
ical and environmental Klebsiella pneumoniae. Genome Med 9:6. https:// review. J Clin Diagn Res 8:DD03–DD04. https://doi.org/10.7860/JCDR/
doi.org/10.1186/s13073-017-0397-1. 2014/9763.5260.
28. Breathnach AS, Cubbon MD, Karunaharan RN, Pope CF, Planche TD. 45. Keren Y, Keshet D, Eidelman M, Geffen Y, Raz-Pasteur A, Hussein K. 2014.
2012. Multidrug-resistant Pseudomonas aeruginosa outbreaks in two Is Leclercia adecarboxylata a new and unfamiliar marine pathogen? J Clin
hospitals: association with contaminated hospital waste-water systems. Microbiol 52:1775–1776. https://doi.org/10.1128/JCM.03239-13.
J Hosp Infect 82:19 –24. https://doi.org/10.1016/j.jhin.2012.06.007. 46. Riazzo C, López-Cerero L, Rojo-Martín MD, Hoyos-Mallecot Y, Fernández-
29. Maheshwari M, Yaser NH, Naz S, Fatima M, Ahmad I. 2016. Emergence of Cuenca F, Martín-Ruíz JL, Pascual-Hernández Á, Naas T, Navarro-Marí JM.
ciprofloxacin-resistant extended-spectrum beta-lactamase-producing 2017. First report of NDM-1-producing clinical isolate of Leclercia adecar-
enteric bacteria in hospital wastewater and clinical sources. J Glob boxylata in Spain. Diagn Microbiol Infect Dis 88:268 –270. https://doi.org/
Antimicrob Resist 5:22–25. https://doi.org/10.1016/j.jgar.2016.01.008. 10.1016/j.diagmicrobio.2017.04.013.
30. Röderová M, Sedláková MH, Pudová V, Hricová K, Silová R, Imwensi PE, 47. Prakash MR, Ravikumar R, Patra N, Indiradevi B. 2015. Hospital-acquired
Bardonˇ J, Kolárˇ M. 2016. Occurrence of bacteria producing broad- pneumonia due to Leclercia adecarboxylata in a neurosurgical centre. J
spectrum beta-lactamases and qnr genes in hospital and urban waste- Postgrad Med 61:123–125. https://doi.org/10.4103/0022-3859.153108.
water samples. New Microbiol 39:124 –133. 48. Treangen TJ, Ondov BD, Koren S, Phillippy AM. 2014. The Harvest suite
31. Yang CM, Lin MF, Liao PC, Yeh HW, Chang BV, Tang TK, Cheng C, Sung for rapid core-genome alignment and visualization of thousands of
CH, Liou ML. 2009. Comparison of antimicrobial resistance patterns intraspecific microbial genomes. Genome Biol 15:524. https://doi.org/10
between clinical and sewage isolates in a regional hospital in Taiwan. .1186/PREACCEPT-2573980311437212.
49. Hoffmann H, Roggenkamp A. 2003. Population genetics of the nomen-
January/February 2018 Volume 9 Issue 1 e02011-17
mbio.asm.org 15

Weingarten et al. ® Downloaded from mbio.asm.org on February 10, 2018 - Published by mbio.asm.org

species Enterobacter cloacae. Appl Environ Microbiol 69:5306 –5318. 61. Kotay S, Chai W, Guilford W, Barry K, Mathers AJ. 2017. Spread from the
https://doi.org/10.1128/AEM.69.9.5306-5318.2003. sink to the patient: in situ study using green fluorescent protein (GFP)-
50. Kizny Gordon AE, Mathers AJ, Cheong EYL, Gottlieb T, Kotay S, Walker expressing Escherichia coli to model bacterial dispersion from hand-
AS, Peto TEA, Crook DW, Stoesser N. 2017. The hospital water environ- washing sink-trap reservoirs. Appl Environ Microbiol 83:e03327-16.
ment as a reservoir for carbapenem-resistant organisms causing https://doi.org/10.1128/AEM.03327-16.
hospital-acquired infections—a systematic review of the literature. Clin
Infect Dis 64:1435–1444. https://doi.org/10.1093/cid/cix132. 62. CLSI. 2017. Performance standards for antimicrobial susceptibility test-
51. Marathe NP, Regina VR, Walujkar SA, Charan SS, Moore ER, Larsson DG, ing, 27th ed. CLSI supplement M100. Clinical and Laboratory Standards
Shouche YS. 2013. A treatment plant receiving waste water from mul- Institute, Wayne, PA.
tiple bulk drug manufacturers is a reservoir for highly multi-drug resis-
tant integron-bearing bacteria. PLoS One 8:e77310. https://doi.org/10 63. CLSI. 2004. Quality control for commercially prepared microbiological
.1371/journal.pone.0077310. culture media; approved standard, 3rd ed. CLSI document M22-A3.
52. Kalaiselvi K, Mangayarkarasi V, Balakrishnan D, Chitraleka V. 2016. Sur- Clinical and Laboratory Standards Institute, Wayne, PA.
vival of antibacterial resistance microbes in hospital-generated recycled
wastewater. J Water Health 14:942–949. https://doi.org/10.2166/wh 64. Conlan S, Lau AF, NISC Comparative Sequencing Program, Palmore
.2016.154. TN, Frank KM, Segre JA. 2016. Complete genome sequence of a
53. Akiba M, Sekizuka T, Yamashita A, Kuroda M, Fujii Y, Murata M, Lee K, Klebsiella pneumoniae strain carrying blaNDM-1 on a multidrug resis-
Joshua DI, Balakrishna K, Bairy I, Subramanian K, Krishnan P, Mu- tance plasmid. Genome Announc 4:e00664-16. https://doi.org/10.1128/
nuswamy N, Sinha RK, Iwata T, Kusumoto M, Guruge KS. 2016. Distribu- genomeA.00664-16.
tion and relationships of antimicrobial resistance determinants among
extended-spectrum-cephalosporin-resistant or carbapenem-resistant 65. Dotson GA, NISC Comparative Sequencing Program, Dekker JP, Palmore
Escherichia coli isolates from rivers and sewage treatment plants in India. TN, Segre JA, Conlan S. 2016. Draft genome sequence of a Klebsiella
Antimicrob Agents Chemother 60:2972–2980. https://doi.org/10.1128/ pneumoniae carbapenemase-positive sequence type 111 Pseudomonas
AAC.01950-15. aeruginosa strain. Genome Announc 4:e01663-15. https://doi.org/10
54. Kochar S, Sheard T, Sharma R, Hui A, Tolentino E, Allen G, Landman D, .1128/genomeA.01663-15.
Bratu S, Augenbraun M, Quale J. 2009. Success of an infection control
program to reduce the spread of carbapenem-resistant Klebsiella pneu- 66. Bankevich A, Nurk S, Antipov D, Gurevich AA, Dvorkin M, Kulikov AS,
moniae. Infect Control Hosp Epidemiol 30:447– 452. https://doi.org/10 Lesin VM, Nikolenko SI, Pham S, Prjibelski AD, Pyshkin AV, Sirotkin AV,
.1086/596734. Vyahhi N, Tesler G, Alekseyev MA, Pevzner PA. 2012. SPAdes: a new
55. Yomoda S, Okubo T, Takahashi A, Murakami M, Iyobe S. 2003. Presence genome assembly algorithm and its applications to single-cell sequenc-
of Pseudomonas putida strains harboring plasmids bearing the metallo- ing. J Comput Biol 19:455– 477. https://doi.org/10.1089/cmb.2012.0021.
beta-lactamase gene bla(IMP) in a hospital in Japan. J Clin Microbiol
41:4246 – 4251. https://doi.org/10.1128/JCM.41.9.4246-4251.2003. 67. Walker BJ, Abeel T, Shea T, Priest M, Abouelliel A, Sakthikumar S, Cuomo
56. Shams AM, Rose LJ, Edwards JR, Cali S, Harris AD, Jacob JT, LaFae A, CA, Zeng Q, Wortman J, Young SK, Earl AM. 2014. Pilon: an integrated
Pineles LL, Thom KA, McDonald LC, Arduino MJ, Noble-Wang JA. 2016. tool for comprehensive microbial variant detection and genome assem-
Assessment of the overall and multidrug-resistant organism bioburden bly improvement. PLoS One 9:e112963. https://doi.org/10.1371/journal
on environmental surfaces in healthcare facilities. Infect Control Hosp .pone.0112963.
Epidemiol 37:1426 –1432. https://doi.org/10.1017/ice.2016.198.
57. Wellington EM, Boxall AB, Cross P, Feil EJ, Gaze WH, Hawkey PM, 68. Assefa S, Keane TM, Otto TD, Newbold C, Berriman M. 2009. ABACAS:
Johnson-Rollings AS, Jones DL, Lee NM, Otten W, Thomas CM, Williams algorithm-based automatic contiguation of assembled sequences. Bioin-
AP. 2013. The role of the natural environment in the emergence of formatics 25:1968 –1969. https://doi.org/10.1093/bioinformatics/btp347.
antibiotic resistance in Gram-negative bacteria. Lancet Infect Dis 13:
155–165. https://doi.org/10.1016/S1473-3099(12)70317-1. 69. Koren S, Walenz BP, Berlin K, Miller JR, Bergman NH, Phillippy AM. 2017.
58. Sheppard AE, Stoesser N, Wilson DJ, Sebra R, Kasarskis A, Anson LW, Canu: scalable and accurate long-read assembly via adaptive k-mer
Giess A, Pankhurst LJ, Vaughan A, Grim CJ, Cox HL, Yeh AJ, Modernising weighting and repeat separation. Genome Res 27:722–736. https://doi
Medical Microbiology (MMM) Informatics Group, Sifri CD, Walker AS, .org/10.1101/gr.215087.116.
Peto TE, Crook DW, Mathers AJ. 2016. Nested Russian doll-like genetic
mobility drives rapid dissemination of the carbapenem resistance gene 70. Krumsiek J, Arnold R, Rattei T. 2007. Gepard: a rapid and sensitive tool for
blaKPC. Antimicrob Agents Chemother 60:3767–3778. https://doi.org/10 creating dotplots on genome scale. Bioinformatics 23:1026 –1028.
.1128/AAC.00464-16. https://doi.org/10.1093/bioinformatics/btm039.
59. Brown CJ, Sen D, Yano H, Bauer ML, Rogers LM, Van der Auwera GA, Top
EM. 2013. Diverse broad-host-range plasmids from freshwater carry few 71. Teer JK, Bonnycastle LL, Chines PS, Hansen NF, Aoyama N, Swift AJ,
accessory genes. Appl Environ Microbiol 79:7684 –7695. https://doi.org/ Abaan HO, Albert TJ, NISC Comparative Sequencing Program, Margulies
10.1128/AEM.02252-13. EH, Green ED, Collins FS, Mullikin JC, Biesecker LG. 2010. Systematic
60. Tofteland S, Naseer U, Lislevand JH, Sundsfjord A, Samuelsen O. 2013. A comparison of three genomic enrichment methods for massively parallel
long-term low-frequency hospital outbreak of KPC-producing Klebsiella DNA sequencing. Genome Res 20:1420 –1431. https://doi.org/10.1101/
pneumoniae involving intergenus plasmid diffusion and a persisting gr.106716.110.
environmental reservoir. PLoS One 8:e59015. https://doi.org/10.1371/
journal.pone.0059015. 72. Thorvaldsdóttir H, Robinson JT, Mesirov JP. 2013. Integrative Genomics
Viewer (IGV): high-performance genomics data visualization and explo-
ration. Brief Bioinform 14:178 –192. https://doi.org/10.1093/bib/bbs017.

73. Ondov BD, Treangen TJ, Melsted P, Mallonee AB, Bergman NH, Koren S,
Phillippy AM. 2016. Mash: fast genome and metagenome distance esti-
mation using MinHash. Genome Biol 17:132. https://doi.org/10.1186/
s13059-016-0997-x.

74. Kurtz S, Phillippy A, Delcher AL, Smoot M, Shumway M, Antonescu C,
Salzberg SL. 2004. Versatile and open software for comparing large
genomes. Genome Biol 5:R12. https://doi.org/10.1186/gb-2004-5-2-r12.

January/February 2018 Volume 9 Issue 1 e02011-17 mbio.asm.org 16

PUBLIC AND ENVIRONMENTAL
HEALTH MICROBIOLOGY

crossm

Spread from the Sink to the Patient: In Downloaded from http://aem.asm.org/ on November 17, 2020 by guest
Situ Study Using Green Fluorescent
Protein (GFP)-Expressing Escherichia coli
To Model Bacterial Dispersion from
Hand-Washing Sink-Trap Reservoirs

Shireen Kotay,a Weidong Chai,a William Guilford,b Katie Barry,a Amy J. Mathersa,c Received 8 December 2016 Accepted 7
February 2017
Division of Infectious Diseases and International Health, Department of Medicine, University of Virginia Health Accepted manuscript posted online 24
System, Charlottesville, Virginia, USAa; Department of Biomedical Engineering, University of Virginia, February 2017
Charlottesville, Virginia, USAb; Clinical Microbiology, Department of Pathology, University of Virginia Health Citation Kotay S, Chai W, Guilford W, Barry K,
System, Charlottesville, Virginia, USAc Mathers AJ. 2017. Spread from the sink to the
patient: in situ study using green fluorescent
ABSTRACT There have been an increasing number of reports implicating Gamma- protein (GFP)-expressing Escherichia coli to
proteobacteria as often carrying genes of drug resistance from colonized sink traps model bacterial dispersion from hand-washing
to vulnerable hospitalized patients. However, the mechanism of transmission from sink-trap reservoirs. Appl Environ Microbiol
the wastewater of the sink P-trap to patients remains poorly understood. Herein we 83:e03327-16. https://doi.org/10.1128/
report the use of a designated hand-washing sink lab gallery to model dispersion of AEM.03327-16.
green fluorescent protein (GFP)-expressing Escherichia coli from sink wastewater to Editor Andrew J. McBain, University of
the surrounding environment. We found no dispersion of GFP-expressing E. coli di- Manchester
rectly from the P-trap to the sink basin or surrounding countertop with coincident Copyright © 2017 Kotay et al. This is an open-
water flow from a faucet. However, when the GFP-expressing E. coli cells were al- access article distributed under the terms of
lowed to mature in the P-trap under conditions similar to those in a hospital envi- the Creative Commons Attribution 4.0
ronment, a GFP-expressing E. coli-containing putative biofilm extended upward over International license.
7 days to reach the strainer. This subsequently resulted in droplet dispersion to the Address correspondence to Amy J. Mathers,
surrounding areas (Ͻ30 in.) during faucet operation. We also demonstrated that [email protected].
P-trap colonization could occur by retrograde transmission along a common pipe. S.K. and W.C. contributed equally to this article.
We postulate that the organisms mobilize up to the strainer from the P-trap, result-
ing in droplet dispersion rather than dispersion directly from the P-trap. This work aem.asm.org 1
helps to further define the mode of transmission of bacteria from a P-trap reservoir
to a vulnerable hospitalized patient.

IMPORTANCE Many recent reports demonstrate that sink drain pipes become
colonized with highly consequential multidrug-resistant bacteria, which then re-
sults in hospital-acquired infections. However, the mechanism of dispersal of
bacteria from the sink to patients has not been fully elucidated. Through estab-
lishment of a unique sink gallery, this work found that a staged mode of trans-
mission involving biofilm growth from the lower pipe to the sink strainer and
subsequent splatter to the bowl and surrounding area occurs rather than splat-
ter directly from the water in the lower pipe. We have also demonstrated that
bacterial transmission can occur via connections in wastewater plumbing to
neighboring sinks. This work helps to more clearly define the mechanism and
risk of transmission from a wastewater source to hospitalized patients in a world
with increasingly antibiotic-resistant bacteria that can thrive in wastewater envi-
ronments and cause infections in vulnerable patients.

KEYWORDS antibiotic resistance, biofilms, dispersion, hand washing, hospital
infections, premise plumbing, sink, transmission, wastewater

April 2017 Volume 83 Issue 8 e03327-16 Applied and Environmental Microbiology

Kotay et al. Applied and Environmental Microbiology Downloaded from http://aem.asm.org/ on November 17, 2020 by guest
aem.asm.org 2
Despite early reports (1–5), the premise that hand-wash sink traps can act as
reservoirs of bacteria that cause nosocomial infections has been frequently over-
looked. There has recently been an alarming increase in sink-related outbreaks world-
wide, with many reports establishing an observational link (6–13). A sink often operates
as an open conduit to wastewater in a patient care area that is often in the same room
as the patient.

Health care establishments often invest in desperate interventions to deal with
nosocomial outbreaks. The preferred method for addressing most of the environment-
related transmission is to employ enhanced cleaning using chemical and physical
agents (14, 15). Unfortunately, routine approaches are inefficient in completely elimi-
nating drug-resistant Gammaproteobacteria in an inaccessible microbiologically active
area such as a sink trap (6, 16–20). The wet, humid, and relatively protected environ-
ment in a sink trap favors the formation of rich stable microbial communities (16, 21,
22). These communities will be exposed to liquids and waste that are discarded in a sink
and may include antimicrobials, discarded beverages, soap, presumably pathogenic
bacteria from health care workers’ hands, and other items. In short, sink traps could
serve as a breeding ground for opportunistic and highly antimicrobial-resistant bacteria
that cannot be easily cleaned or removed (23–28).

There are many reports of a genetic association between pathogens found in sink
traps and those found in patients (29, 30). However, surprisingly little work has been
done to understand the microscale transmission dynamics. It was previously demon-
strated using a suspension of fluorescent particles (Glo Germ; Glo Germ Co., Moab, UT)
that material injected into the P-trap gets dispersed around a hand-washing sink (6).
This result however has not been replicated hitherto in the follow-up studies. Disper-
sion has never been investigated with living organisms. Ultimately, many details remain
unaddressed surrounding the spread of Enterobacteriaceae in sink-trap wastewater
systems: (i) can organisms grow retrograde from the P-trap water to the sink strainer,
(ii) can organisms spread from one sink to another along the internal surfaces of pipes
with shared drainage systems, and (iii) which portion of a colonized drain pipe results
in dispersion into the sink bowl during a hand-washing event? We aim to better
understand the dispersion dynamics of Gammaproteobacteria living in the wastewater
of a sink strainer and P-trap into an area where patients and health care workers could
be exposed. To study this dynamic, we used a surrogate organism that could be easily
tracked while remaining in the Enterobacteriaceae family, where some of the most
concerning threats in antimicrobial resistance are developing (30).

RESULTS
Growth and colonization of GFP-expressing E. coli in the P-trap. In the first 14

days following the installation of the P-trap with established green fluorescent protein
(GFP)-expressing Escherichia coli and just water running from the faucet, GFP-
expressing E. coli was not detected in the tailpipe beyond 1.5 in. above the liquid level
in the P-trap. GFP-expressing E. coli, however, was found to be viable in the P-trap
without any nutrients added. A nutrient regimen was then instituted to understand the
influence of nutrients on mobility and upward growth. The addition of tryptic soy broth
(TSB) promoted GFP-expressing E. coli growth as early as day 1, with growth observed
in the tailpipe 2 in. above the liquid surface in the P-trap (Table 1). On day 7, the strainer
(ϳ8 in. above the liquid in the P-trap) was found to be colonized with GFP-expressing
E. coli. This translates to an average growth rate of 1 in./day along the length of the
tailpipe with the addition of nutrients and without faucet operation. GFP-expressing E.
coli was not detected in the faucet water.

Sink-to-sink transmission of bacteria. In these experiments, a flanking sink (sink 5)
was the only P-trap inoculated with GFP-expressing E. coli and therefore was the sole
source for transmission to the connected sinks. Starting with a lower inoculum con-
centration (103 CFU/ml) in sink 5, on day 7, GFP-expressing E. coli was detected in the
sink 2 and sink 3 P-traps (Fig. 1a). With inoculum concentrations of 106 CFU/ml and
Ͼ1010 CFU/ml in sink 5, all of the sink P-traps in the sink gallery with the exception of

April 2017 Volume 83 Issue 8 e03327-16

GFP-Expressing E. coli Sink-Trap Dispersion Applied and Environmental Microbiology
aem.asm.org 3
TABLE 1 Growth in the tailpipe connected to the P-trap colonized with GFP-expressing
E. coli biofilm

Presence of GFP-expressing E. coli on daya:

Sampling area 01234567

Strainer 8 in. above P-trap water ؊ ؊ ؊ ؊ ؊ ؊ ؊ ؉

Tailpipe ؊؊؊؊؉؉؉؉
6 in. above P-trap water ؊؊؊؉؉؉؉؉
4 in. above P-trap water ؊؉؉؉؉؉؉؉
2 in. above P-trap water

P-trap ؉؉؉؉؉؉؉؉

a“Ϫ” and “ϩ” denote the absence and presence of GFP-expressing E. coli, respectively. Downloaded from http://aem.asm.org/ on November 17, 2020 by guest

sink 1 were found to be colonized with GFP-expressing E. coli after 7 days (Fig. 1b and
c). Faucet water and aerators tested negative for GFP-expressing E. coli. Irrespective of
the starting inoculum concentration, on day 7 the highest level of colonization was
recorded in the sink 3 P-trap. After day 7, when the nutrient regimen (described
previously) was followed for an additional 7 days in each of the sinks in the sink gallery
with an inoculum concentration of Ͼ1010 CFU/ml, GFP-expressing E. coli was detected
in the strainers of sinks 2 and 3 on day 14. This finding validated the upward growth
and growth rate in the tailpipe when nutrients were added. Nonfluorescent colonies
were occasionally observed in the P-trap water samples collected from the sinks, which
were subsequently identified to be Pseudomonas sp. or Stenotrophomonas maltophilia,
and fluorescent colonies were confirmed to be E. coli.

Dispersion of microspheres from sinks. In the first dispersion experiment, when
fluorescent microspheres were inoculated into the offset drain tailpiece only 4 in. below
the strainer, no microspheres were detected on the polyester sheets placed on the
counter space.

However, when the sink bowl was coated with the microspheres, polyester sheets
overlaid on the counter space captured the dispersed microspheres caused by the
faucet operation. Dispersion was observed on almost all zones of the sink counter space
(Fig. 2). Relatively higher levels of dispersion were observed along the major and minor
axes of the elliptical sink bowl (zones 2, 5, 6, 9, 11, and 12). Anterior corners of the sink
counter space (zones 4 and 7), which were most distant from the impact of water in the
sink bowl, received the lowest dispersion.

Dispersion of GFP-expressing E. coli from sinks. Initially the P-trap alone was
inoculated with GFP-expressing E. coli and carefully installed, keeping the tailpipe and
strainer free of GFP-expressing E. coli before operating the faucets. No fluorescent CFU
were observed on the plates placed on the counter or attached to the bowl surface
after faucet operation. Similarly, no fluorescent CFU were detected when GFP-
expressing E. coli was inoculated into the offset drain tailpiece only 4 in. below the
strainer. Interestingly, when there was conspicuous water backup over the strainer as
a result of a higher water flow rate from the faucet than the drainage rate from the
P-trap, dispersal was detected on the plates attached to the bowl surface.

The dispersion pattern recorded when the sink bowl was coated with GFP-
expressing E. coli was comparable to the pattern recorded when the sink bowl was
coated with fluorescent microspheres (Fig. 2). Dispersion was significantly higher along
the axes (zones 6, 9, 11, and 12) and lower at the corners of the sink counter space
(zones 4, 7, and 10).

In contrast, dispersion of GFP-expressing E. coli caused by the faucet operation was
much more extensive when the strainer was allowed to be colonized with GFP-
expressing E. coli prior to the dispersion experiment. In addition to the sink counter
space, we measured dispersion to the sink bowl, faucet, faucet handles, splatter shields,
and the extended counter surface. Dispersion of GFP-expressing E. coli was highest on
the plates attached to the sink bowl (Fig. 3b). Further, dispersion was greater along the

April 2017 Volume 83 Issue 8 e03327-16

Kotay et al. Applied and Environmental Microbiology

Downloaded from http://aem.asm.org/ on November 17, 2020 by guest

FIG 1 GFP-expressing E. coli detected in the P-traps attached to each of the sinks on day 0 (black bars) aem.asm.org 4
and day 7 (gray bars) using (a) 103, (b) 106, and (c) 1010 CFU/ml as the starting inoculum concentrations
in sink 5.

minor axis of the sink bowl (Fig. 3b, zones B3, B4, and B10) than along the major axis
of the sink bowl, associated with a shorter distance from the strike point of the faucet
water to the bowl along this axis. The next highest CFU count from the dispersal was
recorded on the counter area near the faucets (Fig. 3a, zones 12 and 11). A similar
pattern of higher dispersion near the faucets and lower dispersion at the corners of the

April 2017 Volume 83 Issue 8 e03327-16

GFP-Expressing E. coli Sink-Trap Dispersion Applied and Environmental Microbiology

FIG 2 Dispersion of microspheres (gray bars) and GFP-expressing E. coli (black bars) on the area Downloaded from http://aem.asm.org/ on November 17, 2020 by guest
surrounding the sink when the sink bowl was coated. The x axis represents the designated zones of the
sink counter. aem.asm.org 5

counter space (Fig. 3a, zones 4, 7, and 10) was also observed using microspheres.
Dispersion was also recorded in other zones of the counter space, on the Plexiglas
splatter shields, faucets, faucet handles, and extended surface (Fig. 3c). There were no
GFP-expressing E. coli CFU recorded on plates placed beyond 30 in. from the strainer,
demarcating the range of dispersion under these experimental conditions.

Table 2 gives a summary of the total distribution loads recorded using fluorescent
microspheres and GFP-expressing E. coli across each experiment. The loads of disper-
sion on the sink counter were comparable when the sink bowl was coated with
microspheres or GFP-expressing E. coli before the faucet operation. Although the
dispersion load on the sink counter was lower when the sink strainer was colonized, it
is interesting to note that the sink bowl received the highest dispersion.

DISCUSSION
To mimic dispersion in a hospital setting, we first investigated whether GFP-

expressing E. coli would establish consistent colonization in a sink trap as many other
Gammaproteobacteria implicated in nosocomial outbreaks have done (6, 28). Many
recent reports demonstrate that P-traps become colonized with highly consequential
Gammaproteobacteria, which then results in nosocomial transmission (29, 31, 32). The
retained water in a sink P-trap is present to provide a water barrier to prevent
off-gassing of sewer smell, but it may inadvertently provide favorable conditions for
pathogenic and opportunistic antibiotic-resistant microorganisms to survive and de-
velop resilient biofilms (3, 33). However, the mechanism of dispersal of the bacteria in
the P-trap to patients or the surrounding health care area had not been fully elucidated.
We began with the hypothesis that the bacteria originate from the P-trap via droplet
creation when the water from the faucet hits the P-trap water, thus contaminating the
sink bowl and the surrounding area. The finding supporting this theory had been
previously reported using Glo Germ particles (6). However, in the present study with
careful attention to avoid strainer and tail piece contamination, the dispersal directly
from the sink P-trap with either microspheres or GFP-expressing E. coli could not be
reproduced as previously reported (6).

Rather this work demonstrates a different, more staged mode of transmission from
a P-trap reservoir to the sink and surrounding environment. GFP-expressing E. coli in
the P-trap alone was sustained for 14 days but did not grow or mobilize up the tailpipe
to the strainer with just intermittent water exposure. However, when nutrients were

April 2017 Volume 83 Issue 8 e03327-16

Kotay et al. Applied and Environmental Microbiology

Downloaded from http://aem.asm.org/ on November 17, 2020 by guest

FIG 3 Dispersion of GFP-expressing E. coli on the area surrounding the sink when the strainer, tailpipe, aem.asm.org 6
and P-trap were colonized. (a) Sink counter; (b) sink bowl; (c) other surrounding areas. The x axis
represents the designated zones of the sink counter.

April 2017 Volume 83 Issue 8 e03327-16

GFP-Expressing E. coli Sink-Trap Dispersion Applied and Environmental Microbiology

TABLE 2 Comparison of dispersion loads across different experiments

Dispersion load

Dispersion expt Sink counter Sink bowl Faucets and Splatter shields
(>30 in.) faucet handles
Microspheres (microspheres/cm2) NAa NA
Offset drain inoculated with microspheres 0 NA NA NA
Sink bowl coated with microspheres 206 Ϯ 10 NA
0 0
GFP-expressing E. coli (CFU/cm2) 0 NA 0 NA Downloaded from http://aem.asm.org/ on November 17, 2020 by guest
P-trap inoculated with GFP-expressing E. coli NA NA NA
Offset drain inoculated with GFP-expressing E. coli 0 342 Ϯ 17 NA 3Ϯ1
Sink bowl coated with GFP-expressing E. coli 232 Ϯ 17 17 Ϯ 3
Strainer colonized with GFP-expressing E. coli 171 Ϯ 15

aNA, not applicable.

subsequently added to the system, the organisms rapidly grew up the tailpipe to the aem.asm.org 7
strainer at approximately an inch per day. In a real-world setting, motility of bacteria
inside the tailpipe is restricted to relatively sporadic and brief wetting events in which
swimming is an opportunity to colonize new surfaces. It is assumed that once estab-
lished, the biofilm promotes the upward growth of GFP-expressing E. coli in the tailpipe
at an accelerated rate. The nutrient regimen that promoted colonization in our model
reflects our and others’ observations of items commonly disposed of in hospital sinks
(intravenous fluids, feeding supplements, and leftover beverages) (5, 32).

Transmission of bacteria between sinks via a common pipe was a key finding in this
study as this highlights the concept that premise plumbing may be a more continuous
system with shared microbiology than a single isolated sink. The sink gallery used in
this study provided a unique in situ advantage to investigate sink-to-sink transmission
of bacteria through common drains. The two possible mechanisms for P-trap strainers
becoming colonized are seeding of organisms from above and retrograde spread of
organisms along common pipes in a hospital wastewater infrastructure. Here we
demonstrate that it is possible for GFP-expressing E. coli to contaminate adjacent
P-traps with just time and water given a standard U.S. code piping rise of 0.25 ft/ft.
Sink-to-sink or retrograde transmission may explain the recurrence of pathogen colo-
nization following intervention strategies like disinfection or replacement of plumbing
(23). Sink 3 was lowest on the slope in the drain line (see Fig. 4) with arguably the most
opportunity for reflux and retrograde wetting. Sink 1, on the other hand, was farthest
away from the source (sink 5), and its P-trap had the greatest incline in the drain line
connecting the sinks, which could perhaps contribute to the reasons there was no
GFP-expressing E. coli colonization detected in it after 7 days. There has been more
investigation about microbiologic dynamics of infectious viral particles such as those of
severe acute respiratory syndrome (SARS) and Ebola viruses through premise plumbing
systems (34–36). However, the microbiology, sustainability, and dynamics might be
very different, although the backflow and inoculation issues could have some parallels
when comparing viruses to bacteria. As Enterobacteriaceae can either multiply or
remain viable for long periods of time in biofilms coating the interior of P-traps and the
connected plumbing, it may not be sustainable to target any intervention limited to a
single isolated sink as a source of a particular pathogen.

Data from different dispersal experiments suggest that although P-traps can act as
the source or the reservoir of pathogens, the physical presence of the organism in the
sink bowl or colonization of the strainer is necessary for the dispersal to occur.
Colonization of strainers or drains reported in earlier studies (7, 10, 13, 24, 37) was
perhaps a result of ascending biofilm growth from the P-trap to the strainer or
introduction through contaminated fluids. Many of the studies used swab samples,
which likely sampled the strainer rather than P-trap water (17, 20). Once the strainer
was colonized, the water from the faucet resulted in GFP-expressing E. coli dispersion
in the bowl and to the surrounding surfaces of up to 30 in. The range of dispersal

April 2017 Volume 83 Issue 8 e03327-16

Kotay et al. Applied and Environmental Microbiology

FIG 4 Layout of the sink gallery comprising the 5 sink modules and the associated plumbing. Downloaded from http://aem.asm.org/ on November 17, 2020 by guest

recorded in this study was comparable to that reported earlier (6). Greater dispersal aem.asm.org 8
near the faucet may be attributed to the specific designs of the sink bowl and faucet
in this study, which determine the contact angle of water impact. This is an important
finding since many sinks in hospitals are similar in design, with faucet handles repre-
senting a high-touch surface for the sink users (38). It can also be concluded from the
dispersion experiments that secondary and successive dispersals would likely increase
the degree and the scope of dispersion.

There are several limitations to this work. First the use of similar sink bowls across these
sinks only examines the dispersion pattern of this particular sink design. Similarly the
sink-to-sink transmission may not be applicable to all wastewater plumbing systems as the
fixtures on the pipe are very close together, unlike most layouts in health care settings.
However, we speculate that transmission could occur on larger systems over greater
time scales, especially if heavy nutrient and contamination loads were also included.
GFP-expressing E. coli is a laboratory surrogate, and the putative biofilms established in
the short time frame of our experiments are unlikely to be as complex or stable as
biofilms developed in a hospital wastewater system over many years. However, to
address the monomicrobial dominance of the GFP-expressing E. coli added to the
system, we kept the system open, and other environmental organisms were able to
cocolonize in an attempt to mimic the hospital system. Another limitation was the need
to add nutrients to the drain to ensure rapid and robust colonization. We are not clear
how widespread the practice of disposing of dextrose-containing intravenous fluids or
leftover beverages in the hand-wash sinks is; however, we have observed this practice,
and anecdotally it appears to be relatively common in the United States. We also did
not completely characterize the droplet sizes, nor do we demonstrate air sampling to
understand if the dispersion is only droplet or if there are also aerosols that contain
GFP-expressing E. coli. This would require additional testing and is planned as future
work.

In summary, this work for the first time better models the mechanisms of spread of
multidrug-resistant pathogens arising from the sink drain and infecting patients. Drop-
let dispersion from the P-trap does not happen directly. Rather it is a multistage
process: dispersal originates from the strainer and/or the bowl after growth of the
biofilm up from the microbial reservoir of the P-trap. We also demonstrate sink-to-sink
transmission via a common sanitary pipe. This work could have implications for patient
safety, infection control, and interventions as well as the design of future hospital
plumbing systems to eliminate this mode of transmission to vulnerable hospitalized
patients.

MATERIALS AND METHODS

Sink gallery design. A dedicated sink gallery was set up to simulate hospital hand-washing sinks.
The gallery was comprised of five sink modules assembled next to each other (Fig. 4). The five hand-wash
sink stations were identical in bowl designs and dimensions and were modeled from the most common
intensive care unit hand-washing sink type in the acute care hospital at the University of Virginia Medical
Center. Partitions made of 24-in.-high Plexiglas sheet were installed between the sinks to prevent splatter
and cross contamination. Each sink module was built with Corian integrated sink/countertops without an
overflow and fitted with an 8-in. centerset 2-handle gooseneck faucet (Elkay, Oak Brook, IL). The drain line

April 2017 Volume 83 Issue 8 e03327-16

GFP-Expressing E. coli Sink-Trap Dispersion Applied and Environmental Microbiology

FIG 5 (a) Parts of the sink drain line: 1, faucet and handles; 2, sink counter; 3, strainer; 4, tailpipe; 5, sampling ports; Downloaded from http://aem.asm.org/ on November 17, 2020 by guest
6, trap arm; 7, P-trap. (b and c) Schematic of the nutrient regimen (b) and offset drain tailpiece used for dispersion
experiments (c). aem.asm.org 9

under each sink was comprised of a flat-top fixed strainer (drain size, 2 in. by 3 in.), 17-gauge
(1.47-mm-thick) 8- to 10-in.-long tailpipe, P-trap, and trap arms of 0.25-in. outside diameter (o.d.)
(Dearborn Brass-Oatey, Cleveland, OH). All of the fixtures were made of brass with chrome plating. Each
of the sink P-traps was connected to a 3-in. common cast-iron pipe sloping into a T-joint leading into the
building sanitary line located behind sink 3 (Fig. 4).

Inoculation, growth, and establishment of GFP-expressing E. coli in sink P-traps. For the
GFP-expressing E. coli strain (ATCC 25922GFP), the green fluorescent protein (GFP) gene is contained on
a plasmid that also contains an ampicillin resistance gene. A single isolated colony of GFP-expressing E.
coli grown from a Ϫ80°C stock was inoculated into 5 ml tryptic soy broth (TSB) (Becton, Dickinson and
Company, Sparks, MD) containing 100 ␮g/ml ampicillin (ATCC medium 2855). The inoculum concentra-
tion and method varied for each experiment. For establishment of GFP-expressing E. coli in sink P-traps,
new autoclaved P-traps were filled with 100 ml 0.1ϫ TSB and inoculated with ϳ103 CFU/ml GFP-
expressing E. coli. Following inoculation, both ends of the P-traps were covered with perforated Parafilm
(Bemis, Inc., Oshkosh, WI) and allowed to incubate at room temperature (22 Ϯ 2°C) for 14 days to
facilitate adherent bacterial growth. The medium in the P-trap was decanted and replaced with fresh
0.1ϫ TSB every 48 h. An aliquot of decanted medium and a swab sample from the inner surface of the
P-trap were plated on tryptic soy agar (Becton, Dickinson and Company, Sparks, MD) plates containing
100 ␮g/ml ampicillin (TSA) to monitor the growth of GFP-expressing E. coli in the P-traps. TSA plates were
incubated overnight at 37°C, and CFU fluorescing under UV light were enumerated. All preparatory
culturing of GFP-expressing E. coli took place in a separate room from the sink gallery to avoid
unintentional contamination.

Installation of P-traps colonized with GFP-expressing E.coli. After the 14-day incubation, P-traps
were fastened into the plumbing of the sinks (Fig. 5a). The remainder of the drain line was either
autoclaved (strainer, tailpipe, and trap arms) prior to installation or surface disinfected (sink bowl,
countertop, and faucets) with Caviwipes-1 (Meterx Research, Romulus, MI), maintaining at least 1 min of
contact time. After the P-trap was installed, a daily regimen was followed in which 25 ml of TSB followed
by 25 ml of 0.9% NaCl solution (saline) were added in the ratio 1:3 via the strainer (Fig. 5b) to mimic the
potential nutrient exposure in the hospital.

Sampling and enumeration of GFP-expressing E. coli. To monitor the growth of GFP-expressing
E. coli in the plumbing, sampling ports were drilled along the length of the tailpiece (between the P-trap
and the strainer) and the trap arm (between the P-trap and the common line). These holes were fitted
with size 00 silicone stoppers (Cole-Parmer, Vernon Hills, IL) (Fig. 5a). Sterile cotton swabs (Covidien,
Mansfield, MA) presoaked in saline were inserted through these sampling ports, and samples were
collected by turning the swab in a circular motion on the inner surface (ϳ20 cm2) of the tailpipes. Sample
swabs were pulse-vortexed in 3 ml saline, and serial dilutions were plated on TSA. The strainer, faucet
aerator, and bowl surface were sampled with presoaked swabs and processed as described earlier.

Sink-to-sink transmission of bacteria. To investigate sink-to-sink transmission of bacteria, a distal
sink (sink 5) (Fig. 4) was fitted with a P-trap inoculated with GFP-expressing E. coli. The effects of different
inoculum concentrations of GFP-expressing E. coli—103, 106, and Ͼ1010 CFU/ml (colonized for 14

April 2017 Volume 83 Issue 8 e03327-16

Kotay et al. Applied and Environmental Microbiology

FIG 6 (a) Layout of the zones of the sink counter, bowl, and extension surface designated to monitor Downloaded from http://aem.asm.org/ on November 17, 2020 by guest
droplet dispersion and (b) layout of the TSA plates used for GFP-expressing E. coli droplet dispersion on
the surfaces surrounding the sink. aem.asm.org 10

days)—were investigated. Identification to the species level of fluorescent and nonfluorescent colonies
identified from mixed pipe cultures was performed using a matrix-assisted laser desorption-ionization
(MALDI)–time of flight (MALDI-TOF) mass spectrometer (Vitek-MS; bioMérieux, Durham, NC). The waste-
water paths of sinks 1 to 4 were either autoclaved (strainer, tailpipe, P-traps, and trap arms) prior to
installation or surface disinfected (sink bowl, countertop, and faucets) with Caviwipes-1 (Meterx Research,
Romulus, MI). Faucets on each of the five sinks were turned on simultaneously for 1 min, supplying water
at a flow rate of 8 liters/min, once every 24 h for 7 days. No additional feed to any of the sinks was added
during this period of 7 days. On days 0 and 7, P-traps on each of the five sinks were unfastened, and swab
samples from the P-trap were collected and processed as described earlier.

Dispersion measured using fluorescent microspheres. Fluoresbrite YO carboxylate microspheres
(Polysciences, Inc.) which had a 1-␮m diameter and maximum excitation and emission of 529 nm and 546
nm, respectively, were used as a tracer in the preliminary experiments to understand droplet dispersion
from the hand-wash sinks.

To test whether microspheres could be dispersed from below the sink strainer, a 1-ml suspension of
microspheres (ϳ1010 particles) was injected through a strainer attached to a Hert 4.5-in. offset drain
tailpiece typically used for wheelchair-accessible sinks (American Standard, model 7723018.002) (Fig. 5c).
The vertical distance between the strainer and microsphere suspension injected into the tailpipe was ϳ4
in. Counter space around the sink bowl was thoroughly wiped with alcohol wipes (Covidien Webcol 6818;
Kendall), and polyester sheets precut to appropriate shapes were placed on the counter to cover the
entire sink counter and labeled according to position (Fig. 6a). The faucet was turned on for 5 min at a
water flow rate of 1.8 to 3.0 liters/min. Polyester sheets were harvested and immediately analyzed using
a ChemiDoc MP system (Bio-Rad Laboratories, Inc.) with an exposure time of 5 s. Fluorescent micro-
spheres were enumerated from the digital micrographs using the Image Lab Software (Bio-Rad Labo-
ratories, Inc.).

To test whether microspheres could be dispersed from the surface of the sink bowl, the sink bowl was
evenly coated with a 20-ml microsphere suspension (ϳ1010 particles/ml) using a disposable swab (Sage
Products, Inc., Cary, IL), and the dispersion experiment was repeated following the protocol described
above. To ascertain there was no nonspecific background fluorescence in the sink and/or the water from
the faucet, a control using the same protocol but without the fluorescent microspheres was performed
before each experiment.

April 2017 Volume 83 Issue 8 e03327-16

GFP-Expressing E. coli Sink-Trap Dispersion Applied and Environmental Microbiology

Dispersion measured using GFP-expressing E.coli. Dispersion using GFP-expressing E. coli was Downloaded from http://aem.asm.org/ on November 17, 2020 by guest
investigated in three experiments. To test whether live organisms in the P-trap could be dispersed by
running water, ϳ1010 CFU/ml GFP-expressing E. coli in saline was added to an autoclaved P-trap and
fitted into the drain line that was preautoclaved (strainer, tailpipe, and trap arms). Similarly, to test
whether live organisms could be dispersed from the tailpieces of wheelchair-accessible sinks, a suspen-
sion of ϳ1010 CFU/ml GFP-expressing E. coli was added with a syringe through the strainer into the Hert
4.5-ft offset drain tailpiece (Fig. 5c). Just as in the microsphere dispersion experiment, the vertical
distance between the strainer and GFP-expressing E. coli suspension injected into the tailpipe was ϳ4 in.

We next tested whether live organisms from the surface of the sink bowl could be dispersed by
running water. The sink bowl surface was evenly coated with an approximately 20-ml suspension of 1010
CFU/ml GFP-expressing E. coli.

Finally, to mimic all of these conditions, a P-trap colonized with GFP-expressing E. coli (for 14 days)
was installed, and a nutrient regimen (Fig. 5b) was followed for 14 days to intentionally promote the
GFP-expressing E. coli colonization in the attached tailpipe and strainer. On day 15, the dispersion
experiment was performed.

Before each of the GFP-expressing E. coli dispersion experiments, the counter space was thoroughly
disinfected with Caviwipes-1. TSA plates were then positioned on the sink counter surrounding the bowl
and an extension platform (Fig. 6b). Additional plates were attached to the sink bowl, faucets, Plexiglas
partitions, and faucet handles using adhesive tape. TSA plates were also placed 3 m away from the sink
as negative controls. The faucet was turned on for 5 min with a water flow rate of 1.8 to 3.0 liters/min.
Lids of the TSA plates were removed only during faucet operation. Swab samples from the faucet
aerators before and after operation were collected and plated on TSA. Prior to each dispersion
experiment, 50 ml water from the faucet was also collected, and aliquots were plated to assess for the
presence of GFP-expressing E. coli in source water and ensure cross contamination of GFP-expressing E.
coli had not occurred. A control dispersion experiment was also performed using the same protocol prior
to GFP-expressing E. coli inoculation in each case. Dispersion per defined area (CFU per square
centimeter) was deduced by dividing the CFU counts in the TSA plate by the surface area of the TSA
plate.

REFERENCES patient environment. J Hosp Infect 85:106 –111. https://doi.org/10.1016/
j.jhin.2013.07.006.
1. Edmonds P, Suskind RR, Macmillan BG, Holder IA. 1972. Epidemiology of 12. Knoester M, de Boer M, Maarleveld J, Claas E, Bernards A, Jonge E, van
Pseudomonas-aeruginosa in a burns hospital—surveillance by a com- Dissel J, Veldkamp K. 2014. An integrated approach to control a pro-
bined typing system. Appl Microbiol 24:219 –225. longed outbreak of multidrug-resistant Pseudomonas aeruginosa in an
intensive care unit. Clin Microbiol Infect 20:O207–O215. https://doi.org/
2. Perryman F, Flournoy D. 1980. Prevalence of gentamicin-resistant and 10.1111/1469-0691.12372.
amikacin-resistant bacteria in sink drains. J Clin Microbiol 12:79 – 83. 13. Leitner E, Zarfel G, Luxner J, Herzog K, Pekard-Amenitsch S, Hoenigl M,
Valentin T, Feierl G, Grisold AJ, Hogenauer C, Sill H, Krause R, Zollner-
3. Levin M, Olson B, Nathan C, Kabins S, Weinstein R. 1984. Pseudomonas Schwetz I. 2015. Contaminated handwashing sinks as the source of a
in the sinks in an intensive-care unit—relation to patients. J Clin Pathol clonal outbreak of KPC-2-producing Klebsiella oxytoca on a hematology
37:424 – 427. https://doi.org/10.1136/jcp.37.4.424. ward. Antimicrob Agents Chemother 59:714 –716. https://doi.org/10
.1128/AAC.04306-14.
4. French G, Shannon K, Simmons N. 1996. Hospital outbreak of Klebsiella 14. Carling PC, Bartley JM. 2010. Evaluating hygienic cleaning in health care
pneumoniae resistant to broad-spectrum cephalosporins and beta- settings: what you do not know can harm your patients. Am J Infect
lactam– beta-lactamase inhibitor combinations by hyperproduction of Control 38:S41–S50. https://doi.org/10.1016/j.ajic.2010.03.004.
SHV-5 beta-lactamase. J Clin Microbiol 34:358 –363. 15. Weber D, Rutala W, Miller M, Huslage K, Sickbert-Bennett E. 2010. Role of
hospital surfaces in the transmission of emerging health care-associated
5. Rutala WA, Weber DJ. 1997. Water as a reservoir of nosocomial patho- pathogens: Norovirus, Clostridium difficile, and Acinetobacter species. Am
gens. Infect Control Hosp Epidemiol 18:609 – 616. J Infect Control 38:S25–S33. https://doi.org/10.1016/j.ajic.2010.04.196.
16. Dancer SJ. 2014. Controlling hospital-acquired infection: focus on the
6. Hota S, Hirji Z, Stockton K, Lemieux C, Dedier H, Wolfaardt G, Gardam M. role of the environment and new technologies for decontamination. Clin
2009. Outbreak of multidrug-resistant Pseudomonas aeruginosa coloni- Microbiol Rev 27:665– 690. https://doi.org/10.1128/CMR.00020-14.
zation and infection secondary to imperfect intensive care unit room 17. Vergara-Lopez S, Dominguez M, Conejo M, Pascual A, Rodriguez-Bano J.
design. Infect Control Hosp Epidemiol 30:25–33. https://doi.org/10.1086/ 2013. Wastewater drainage system as an occult reservoir in a protracted
592700. clonal outbreak due to metallo-beta-lactamase-producing Klebsiella oxy-
toca. Clin Microbiol Infect 19:E490 –E498. https://doi.org/10.1111/1469
7. Hong K, Oh H, Song J, Lim J, Kang D, Son I, Park J, Kim E, Lee H, Choi E. -0691.12288.
2012. Investigation and control of an outbreak of imipenem-resistant 18. Chapuis A, Amoureux L, Bador J, Gavalas A, Siebor E, Chretien ML, Caillot
Acinetobacter baumannii infection in a pediatric intensive care unit. D, Janin M, de Curraize C, Neuwirth C. 2016. Outbreak of extended-
Pediatr Infect Dis J 31:685– 690. https://doi.org/10.1097/INF spectrum beta-lactamase producing Enterobacter cloacae with high MICs
.0b013e318256f3e6. of quaternary ammonium compounds in a hematology ward associated
with contaminated sinks. Front Microbiol 7:1070. https://doi.org/10
8. Lowe C, Willey B, O’Shaughnessy A, Lee W, Lum M, Pike K, Larocque C, .3389/fmicb.2016.01070.
Dedier H, Dales L, Moore C, McGeer A. 2012. Outbreak of extended- 19. Clarivet B, Grau D, Jumas-Bilak E, Jean-Pierre H, Pantel A, Parer S, Lotthe
spectrum beta-lactamase-producing Klebsiella oxytoca infections associ- A. 28 April 2016. Persisting transmission of carbapenemase-producing
ated with contaminated handwashing sinks. Emerg Infect Dis 18: Klebsiella pneumoniae due to an environmental reservoir in a university
1242–1247. https://doi.org/10.3201/eid1808.111268. hospital, France, 2012 to 2014. Euro Surveill https://doi.org/10.2807/
1560-7917.ES.2016.21.17.30213.
9. Starlander G, Melhus A. 2012. Minor outbreak of extended-spectrum 20. Wolf I, Bergervoet P, Sebens F, van den Oever H, Savelkoul P, van der
beta-lactamase-producing Klebsiella pneumoniae in an intensive care
unit due to a contaminated sink. J Hosp Infect 82:122–124. https://doi aem.asm.org 11
.org/10.1016/j.jhin.2012.07.004.

10. Kotsanas D, Wijesooriya W, Korman T, Gillespie E, Wright L, Snook K,
Williams N, Bell J, Li H, Stuart R. 2013. “Down the drain”: carbapenem-
resistant bacteria in intensive care unit patients and handwashing sinks.
Med J Aust 198:267–269. https://doi.org/10.5694/mja12.11757.

11. Roux D, Aubier B, Cochard H, Quentin R, van der Mee-Marquet N. 2013.
Contaminated sinks in intensive care units: an underestimated source of
extended-spectrum beta-lactamase-producing Enterobacteriaceae in the

April 2017 Volume 83 Issue 8 e03327-16

Kotay et al. Applied and Environmental Microbiology

Zwet W. 2014. The sink as a correctable source of extended-spectrum 29. Decker B, Palmore T. 2013. The role of water in healthcare-associated Downloaded from http://aem.asm.org/ on November 17, 2020 by guest
beta-lactamase contamination for patients in the intensive care unit. infections. Curr Opin Infect Dis 26:345–351. https://doi.org/10.1097/QCO
J Hosp Infect 87:126 –130. https://doi.org/10.1016/j.jhin.2014.02.013. .0b013e3283630adf.
21. Vickery K, Deva A, Jacombs A, Allan J, Valente P, Gosbell I. 2012. Presence
of biofilm containing viable multiresistant organisms despite terminal 30. Centers for Disease Control and Prevention. 2013. Antibiotic resistance
cleaning on clinical surfaces in an intensive care unit. J Hosp Infect threats in the United States. Centers for Disease Control and Prevention,
80:52–55. https://doi.org/10.1016/j.jhin.2011.07.007. Office of Infectious Disease, Atlanta, GA.
22. Dancer SJ. 2016. Dos and don’ts for hospital cleaning. Curr Opin Infect
Dis 29:415– 423. https://doi.org/10.1097/QCO.0000000000000289. 31. Ferranti G, Marchesi I, Favale M, Borella P, Bargellini A. 2014. Aetiology,
23. La Forgia C, Franke J, Hacek D, Thomson R, Robicsek A, Peterson L. 2010. source and prevention of waterborne healthcare-associated infections: a
Management of a multidrug-resistant Acinetobacter baumannii outbreak review. J Med Microbiol 63:1247–1259. https://doi.org/10.1099/jmm.0
in an intensive care unit using novel environmental disinfection: a .075713-0.
38-month report. Am J Infect Control 38:259 –263. https://doi.org/10
.1016/j.ajic.2009.07.012. 32. Kanamori H, Weber DJ, Rutala WA. 2016. Healthcare outbreaks associ-
24. Breathnach AS, Cubbon MD, Karunaharan RN, Pope CF, Planche TD. ated with a water reservoir and infection prevention strategies. Clin
2012. Multidrug-resistant Pseudomonas aeruginosa outbreaks in two Infect Dis 62:1423–1435. https://doi.org/10.1093/cid/ciw122.
hospitals: association with contaminated hospital waste-water sys-
tems. J Hosp Infect 82:19 –24. https://doi.org/10.1016/j.jhin.2012.06 33. McBain AJ, Bartolo RG, Catrenich CE, Charbonneau D, Ledder RG, Rickard
.007. AH, Symmons SA, Gilbert P. 2003. Microbial characterization of biofilms
25. Tofteland S, Naseer U, Lislevand JH, Sundsfjord A, Samuelsen O. 2013. A in domestic drains and the establishment of stable biofilm microcosms.
long-term low-frequency hospital outbreak of KPC-producing Klebsiella Appl Environ Microbiol 69:177–185. https://doi.org/10.1128/AEM.69.1
pneumoniae involving intergenus plasmid diffusion and a persisting .177-185.2003.
environmental reservoir. PLoS One 8:e59015. https://doi.org/10.1371/
journal.pone.0059015. 34. Li Y, Duan S, Yu IT, Wong TW. 2005. Multi-zone modeling of probable SARS
26. Stjarne Aspelund A, Sjostrom K, Olsson Liljequist B, Morgelin M, Me- virus transmission by airflow between flats in Block E, Amoy Gardens.
lander E, Pahlman LI. 2016. Acetic acid as a decontamination method for Indoor Air 15:96–111. https://doi.org/10.1111/j.1600-0668.2004.00318.x.
sink drains in a nosocomial outbreak of metallo-beta-lactamase-
producing Pseudomonas aeruginosa. J Hosp Infect 94:13–20. https://doi 35. Ye Y, Ellenberg RM, Graham KE, Wigginton KR. 2016. Survivability, par-
.org/10.1016/j.jhin.2016.05.009. titioning, and recovery of enveloped viruses in untreated municipal
27. Soothill J. 2016. Carbapenemase-bearing Klebsiella spp. in sink drains: wastewater. Environ Sci Technol 50:5077–5085. https://doi.org/10.1021/
investigation into the potential advantage of copper pipes. J Hosp Infect acs.est.6b00876.
93:152–154. https://doi.org/10.1016/j.jhin.2016.03.013.
28. Fusch C, Pogorzelski D, Main C, Meyer C, el Helou S, Mertz D. 2015. 36. Lin K, Marr LC. 10 February 2017. Aerosolization of Ebola virus surrogates
Self-disinfecting sink drains reduce the Pseudomonas aeruginosa biobur- in wastewater systems. Environ Sci Technol https://doi.org/10.1021/acs
den in a neonatal intensive care unit. Acta Paediatr 104:e344 – e349. .est.6b04846.
https://doi.org/10.1111/apa.13005.
37. Ambrogi V, Cavalie L, Mantion B, Ghiglia MJ, Cointault O, Dubois D, Prere
MF, Levitzki N, Kamar N, Malavaud S. 2016. Transmission of metallo-
beta-lactamase-producing Pseudomonas aeruginosa in a nephrology-
transplant intensive care unit with potential link to the environment. J
Hosp Infect 92:27–29. https://doi.org/10.1016/j.jhin.2015.09.007.

38. Devnani M, Kumar R, Sharma RK, Gupta AK. 2011. A survey of hand-
washing facilities in the outpatient department of a tertiary care teach-
ing hospital in India. J Infect Dev Ctries 5:114 –118.

April 2017 Volume 83 Issue 8 e03327-16 aem.asm.org 12

Contaminated Handwashing Sinks as the Source of a Clonal Outbreak
of KPC-2-Producing Klebsiella oxytoca on a Hematology Ward

Eva Leitner,a Gernot Zarfel,a Josefa Luxner,a Kathrin Herzog,b Shiva Pekard-Amenitsch,c Martin Hoenigl,d Thomas Valentin,d
Gebhard Feierl,a Andrea J. Grisold,a Christoph Högenauer,b Heinz Sill,e Robert Krause,d Ines Zollner-Schwetzd
Institute of Hygiene Microbiology and Environmental Medicine, Medical University of Graz, Graz, Austriaa; Division of Gastroenterology and Hepatology, Department of
Internal Medicine, Medical University of Graz, Graz, Austriab; Austrian Agency for Health and Food Safety, Institute of Medical Microbiology and Hygiene, Graz, Austriac;
Section of Infectious Diseases and Tropical Medicine, Department of Internal Medicine, Medical University of Graz, Graz, Austriad; Division of Hematology, Department of
Internal Medicine, Medical University of Graz, Graz, Austriae

We investigated sinks as possible sources of a prolonged Klebsiella pneumonia carbapenemase (KPC)-producing Klebsiella oxy-
toca outbreak. Seven carbapenem-resistant K. oxytoca isolates were identified in sink drains in 4 patient rooms and in the medi-
cation room. Investigations for resistance genes and genetic relatedness of patient and environmental isolates revealed that all
the isolates harbored the blaKPC-2 and blaTEM-1 genes and were genetically indistinguishable. We describe here a clonal outbreak
caused by KPC-2-producing K. oxytoca, and handwashing sinks were a possible reservoir.

The prevalence of carbapenem-resistant Enterobacteriaceae has Screening swabs from dry surfaces are taken and tested rou-
increased over the past 10 years (1). The production of Kleb- tinely 4 times per year by our infection control team and have
siella pneumonia carbapenemases (KPCs) is one way of conferring never yielded K. oxytoca. In addition, in two publications dealing
resistance to carbapenems belonging to Ambler class A (2). Nos- with nosocomial outbreaks of K. oxytoca, swabs from dry surfaces
ocomial infections due to KPC-producing Enterobacteriaceae are in the environment surrounding the patients did not yield any K.
of increasing concern, especially in long-term care facilities and oxytoca isolates. In contrast, K. oxytoca was identified in sinks and
intensive care units (ICUs) (2). In 2011, a nosocomial outbreak of drains (9, 10). Therefore, we chose to focus our investigation on
KPC-producing Klebsiella oxytoca occurred in a medical ICU at sinks. Fifty-eight swabs were taken from handwashing sink drains,
the Medical University of Graz in Graz, Austria. Five patients, all 23 from handwashing sink overflows (not all sinks had overflows),
of whom survived, were affected (3). and 19 from shower drains in the Division of Hematology. Swabs
from sink surfaces were taken from contaminated sinks in a sec-
From October 2011 to October 2013, KPC-producing K. oxy- ond round of testing.
toca isolates with resistance patterns identical to that of the out-
break strain were identified from 10 more patients in the Division Swabs were seeded onto MacConkey agar and chromID Carba
of Hematology ward. As K. oxytoca isolates were detected over a (both from bioMérieux, Marcy l’Étoile, France) and analyzed ac-
2-year period, a reservoir in the environment was suspected. Nos- cording to microbiological standards (12). The isolates were
ocomial outbreaks with K. oxytoca have been reported previously screened for the presence of resistance genes with the DNA mi-
(4–10). In two publications, handwashing sinks, especially sink croarray-based Check-MDR 103 kit (Check-Points, Wageningen,
drains, were identified as the source of the outbreaks (9, 10). Netherlands) according to the manufacturer’s protocol (http:
//www.check-points.com/support/manuals/). Sequencing of the
The aim of this study was to investigate sink drains as possible detected carbapenemases and other beta-lactamase gene families
sources of the KPC-producing K. oxytoca isolates, the genetic re- was performed as previously described (13, 14). Repetitive se-
latedness of patient and environmental isolates, and the underly- quence-based PCR (rep-PCR) was carried out with the DiversiLab
ing mechanism of carbapenem resistance. system (bioMérieux, Nürtingen, Germany). In addition, all the
isolates were tested using multilocus sequence typing (MLST), as
The outbreak occurred at the Division of Hematology at the described previously (15).
Medical University of Graz, a tertiary care facility. This 28-bed
ward has approximately 1,200 admissions annually, almost exclu- Eleven K. oxytoca isolates were found in the sink drains. Seven
sively of patients with hematological malignancies. The index pa-
tient involved in the outbreak of KPC-producing K. oxytoca was in Received 15 September 2014 Returned for modification 13 October 2014
the medical ICU in December 2010 (3). The patient was consid- Accepted 23 October 2014
ered colonized but not infected according to CDC criteria (11).
The patient was then transferred to the hematology ward, where Accepted manuscript posted online 27 October 2014
contact precautions were continued.
Citation Leitner E, Zarfel G, Luxner J, Herzog K, Pekard-Amenitsch S, Hoenigl M,
KPC-producing K. oxytoca isolates with resistance patterns Valentin T, Feierl G, Grisold AJ, Högenauer C, Sill H, Krause R, Zollner-Schwetz I.
identical to those of the outbreak strain were identified from 10 2015. Contaminated handwashing sinks as the source of a clonal outbreak of
more patients in the Division of Hematology. Three patients were KPC-2-producing Klebsiella oxytoca on a hematology ward. Antimicrob Agents
identified in late 2011 and 2012, and seven were detected from Chemother 59:714 –716. doi:10.1128/AAC.04306-14.
January to July 2013. Six patients developed an infection (Table 1).
All the patients were admitted more than once during the out- Address correspondence to Ines Zollner-Schwetz, [email protected].
break period either for chemotherapy or because of side effects
during chemotherapy. Copyright © 2015, American Society for Microbiology. All Rights Reserved.

doi:10.1128/AAC.04306-14

714 aac.asm.org Antimicrobial Agents and Chemotherapy January 2015 Volume 59 Number 1

Prolonged KPC-2-Producing Klebsiella oxytoca Outbreak

a m, male; f, female. 10 56, m 9 8 7 6 5 60, f 4 51, f 3 50, f Patient TABLE 1 Clinical data and epidemiological information isolates showed multidrug resistance, including resistance to car-
b AML, acute myeloid leukemia. no. bapenems, one was an extended-spectrum beta-lactamase
c BAL, bronchoalveolar lavage. (ESBL)-producing K. oxytoca isolate, and three isolates showed
d KO, KPC-2- producing K. oxytoca. 1 wild-type resistance. The 7 KPC-producing K. oxytoca isolates
2 were found in rooms 23, 25, and 29 (double rooms housing 2
patients each with two sinks), room 37 (a single room), and 32A (a
89, m 54, m 72, m 61, m Patient age room in which medication is prepared by the nursing staff). Five
(yr) and KPC-producing K. oxytoca isolates were detected in sink drains (in
gendera rooms 23, 25, 29, 37, and 32A). One isolate was found in the sink
overflow (room 37), and one was found in the shower drain
39, m (room 25). Swabs taken from sink surfaces in a second round did
65, m not yield any additional KPC-producing K. oxytoca isolates.

AML B-cell non-Hodgkin lymphoma AML secondary AML Myelodysplastic syndrome, B-cell non-Hodgkin lymphoma AML AML AML Comorbidity All the isolates from the patients and the seven isolates derived
from sinks were multidrug resistant, exhibited susceptibility to
AMLb amikacin, colistin, and fosfomycin only, and were found to be
AML KPC producers. When microarray technology was used, blaKPC
and blaTEM were detected in all the isolates and were identified as
Date of first blaKPC-2 and blaTEM-1 by sequencing. rep-PCR revealed that all
detection strains tested were indistinguishable with a similarity index of
(day/mo/yr) Ͼ97.5%. MLST yielded only one sequence type, ST4, for all the
25/10/2011 KPC-2-producing K. oxytoca isolates.
28/12/2011
The starting point for this outbreak was a colonized patient
13/11/2012 from the ICU who later was transferred to the hematology ward.
We hypothesize that in the case of this patient, KPC-2-producing
8/1/2013 K. oxytoca got into the sink most likely during personal hygiene
activities or by the disposal of contaminated body fluids where it
15/1/2013 persisted. In the Division of Hematology, the water from the sink
9/4/2013 faucets directly hits the mesh that covers the sink drain. We hy-
14/05/2013 pothesize that some patients were colonized by contaminated
aerosols when using the sinks for personal hygiene; 6 of 10 affected
17/5/2013 patients stayed in rooms with contaminated sinks, and 2 patients
31/7/2013 shared a room with a patient who later proved to be infected or
colonized. We speculate that cross-contamination took place ei-
31/10/2013 ther by direct contact between the patients or through the hands of
health care workers. The potential involvement of the health care
Site(s) of detection workers is indicated by the fact that KPC-2-producing K. oxytoca
Anal swab was found in the sink of the room where medication is prepared by
BALc the health care staff.

Blood, skin, stool, urine, abscess As a consequence of the prolonged outbreak, the existing in-
fection control measures (isolating colonized patients, enforcing
Stool, BAL hand hygiene measures, and cleaning the ward, particularly the
sinks and equipment) were reinforced. An exchange of all the
Stool colonized sinks is under way. Since October 2013, no more KPC-
Anal swab, throat, BAL producing K. oxytoca isolates have been identified.
Stool
In conclusion, a clonal relationship between environmental
BAL, throat, anal swab isolates and patient strains was determined and pointed to hand-
Catheter urine, throat washing sinks as a possible reservoir for the prolonged nosocomial
outbreak of KPC-2-producing K. oxytoca.
Blood, throat, anal swab, skin, stool
REFERENCES
Infection or
colonization 1. Nordmann P, Naas T, Poirel L. 2011. Global spread of carbapenemase-
Colonization producing Enterobacteriaceae. Emerg Infect Dis 17:1791–1798. http://dx
Infection .doi.org/10.3201/eid1710.110655.

Infection 2. Levy Hara G, Gould I, Endimiani A, Pardo PR, Daikos G, Hsueh PR,
Mehtar S, Petrikkos G, Casellas JM, Daciuk L, Paciel D, Novelli A,
Infection Saginur R, Pryluka D, Medina J, Savio E. 2013. Detection, treatment,
and prevention of carbapenemase-producing Enterobacteriaceae: recom-
Colonization mendations from an International Working Group. J Chemother 25:129 –
Infection 140. http://dx.doi.org/10.1179/1973947812Y.0000000062.
Colonization
3. Hoenigl M, Valentin T, Zarfel G, Wuerstl B, Leitner E, Salzer HJ, Posch
Infection J, Krause R, Grisold AJ. 2012. Nosocomial outbreak of Klebsiella pneu-
Colonization moniae carbapenemase-producing Klebsiella oxytoca in Austria. Antimi-

Infection

Outcome
Alive
Death due to KOd

pneumonia
Death due to KO abdominal

wall abscess
Death due to AML relapse

Alive
Death due to KO pneumonia
Alive

Death due to KO pneumonia
Alive

Alive

Patient stayed in a room
with K. oxytoca in sink
(room no.)
No
No

Yes (23, 25)

No, but shared room with
patient 5

Yes (25)
Yes (25)
Yes (23)

Yes (25)
No, but shared room with

patient 8
Yes (25)

January 2015 Volume 59 Number 1 Antimicrobial Agents and Chemotherapy aac.asm.org 715

Leitner et al.

crob Agents Chemother 56:2158 –2161. http://dx.doi.org/10.1128/AAC 10. Vergara-Lopez S, Dominguez MC, Conejo MC, Pascual A, Rodriguez-
.05440-11. Bano J. 2013. Wastewater drainage system as an occult reservoir in a
4. Sardan YC, Zarakolu P, Altun B, Yildirim A, Yildirim G, Hascelik G, protracted clonal outbreak due to metallo-beta-lactamase-producing
Uzun O. 2004. A cluster of nosocomial Klebsiella oxytoca bloodstream Klebsiella oxytoca. Clin Microbiol Infect 19:E490 –E498. http://dx.doi.org
infections in a university hospital. Infect Control Hosp Epidemiol 25: /10.1111/1469-0691.12288.
878 – 882. http://dx.doi.org/10.1086/502313.
5. Jeong SH, Kim WM, Chang CL, Kim JM, Lee K, Chong Y, Hwang HY, 11. Horan TC, Andrus M, Dudeck MA. 2008. CDC/NHSN surveillance
Baek YW, Chung HK, Woo IG, Ku JY. 2001. Neonatal intensive care unit definition of health care-associated infection and criteria for specific types
outbreak caused by a strain of Klebsiella oxytoca resistant to aztreonam due of infections in the acute care setting. Am J Infect Control 36:309 –332.
to overproduction of chromosomal beta-lactamase. J Hosp Infect 48:281– http://dx.doi.org/10.1016/j.ajic.2008.03.002.
288. http://dx.doi.org/10.1053/jhin.2001.1018.
6. Zarate MS, Gales AC, Picao RC, Pujol GS, Lanza A, Smayevsky J. 12. Lee K, Lim YS, Yong D, Yum JH, Chong Y. 2003. Evaluation of the
2008. Outbreak of OXY-2-producing Klebsiella oxytoca in a renal trans- Hodge test and the imipenem-EDTA double-disk synergy test for differ-
plant unit. J Clin Microbiol 46:2099 –2101. http://dx.doi.org/10.1128 entiating metallo-beta-lactamase-producing isolates of Pseudomonas spp.
/JCM.00194-08. and Acinetobacter spp. J Clin Microbiol 41:4623– 4629. http://dx.doi.org
7. Schulz-Stubner S, Kniehl E. 2011. Transmission of extended-spectrum /10.1128/JCM.41.10.4623-4629.2003.
beta-lactamase Klebsiella oxytoca via the breathing circuit of a transport
ventilator: root cause analysis and infection control recommendations. 13. Bradford PA, Bratu S, Urban C, Visalli M, Mariano N, Landman D,
Infect Control Hosp Epidemiol 32:828 – 829. http://dx.doi.org/10.1086 Rahal JJ, Brooks S, Cebular S, Quale J. 2004. Emergence of carbapenem-
/661225. resistant Klebsiella species possessing the class A carbapenem-hydrolyzing
8. Ruiz E, Rezusta A, Saenz Y, Rocha-Gracia R, Vinue L, Vindel A, KPC-2 and inhibitor-resistant TEM-30 beta-lactamases in New York City.
Villuendas C, Azanedo ML, Monforte ML, Revillo MJ, Torres C. 2011. Clin Infect Dis 39:55– 60. http://dx.doi.org/10.1086/421495.
New genetic environments of aac(6=)-Ib-cr gene in a multiresistant Kleb-
siella oxytoca strain causing an outbreak in a pediatric intensive care unit. 14. Zarfel G, Galler H, Feierl G, Haas D, Kittinger C, Leitner E, Grisold
Diagn Microbiol Infect Dis 69:236 –238. http://dx.doi.org/10.1016/j AJ, Mascher F, Posch J, Pertschy B, Marth E, Reinthaler FF. 2013.
.diagmicrobio.2010.09.004. Comparison of extended-spectrum-beta-lactamase (ESBL) carrying
9. Lowe C, Willey B, O’Shaughnessy A, Lee W, Lum M, Pike K, Larocque Escherichia coli from sewage sludge and human urinary tract infection.
C, Dedier H, Dales L, Moore C, McGeer A, Mount Sinai Hospital Environ Pollut 173:192–199. http://dx.doi.org/10.1016/j.envpol.2012.09
Infection Control Team. 2012. Outbreak of extended-spectrum beta- .019.
lactamase-producing Klebsiella oxytoca infections associated with con-
taminated handwashing sinks. Emerg Infect Dis 18:1242–1247. http://dx 15. Herzog KA, Schneditz G, Leitner E, Feierl G, Hoffmann KM, Zollner-
.doi.org/10.3201/eid1808.111268. Schwetz I, Krause R, Gorkiewicz G, Zechner EL, Hogenauer C. 2014.
Genotypes of Klebsiella oxytoca isolates from patients with nosocomial
pneumonia are distinct from those of isolates from patients with antibi-
otic-associated hemorrhagic colitis. J Clin Microbiol 52:1607–1616. http:
//dx.doi.org/10.1128/JCM.03373-13.

716 aac.asm.org Antimicrobial Agents and Chemotherapy January 2015 Volume 59 Number 1

Clinical Infectious Diseases Downloaded from https://academic.oup.com/cid/article/65/6/935/3831257 by guest on 09 May 2022
MAJOR ARTICLE

Control of Multidrug-Resistant Pseudomonas aeruginosa in
Allogeneic Hematopoietic Stem Cell Transplant Recipients
by a Novel Bundle Including Remodeling of Sanitary and
Water Supply Systems

Annelene Kossow,1 Stefanie Kampmeier,1 Stefanie Willems,1 Wolfgang E. Berdel,2 Andreas H. Groll,3 Birgit Burckhardt,3 Claudia Rossig,3
Christoph Groth,2 Evgeny A. Idelevich,4 Frank Kipp,1 Alexander Mellmann,1,a and Matthias Stelljes2,a
1Institute of Hygiene, 2Department of Medicine A, Hematology and Oncology, 3University Children’s Hospital Muenster, Department of Pediatric Hematology and Oncology, and 4Institute of Medical
Microbiology, University of Muenster, Germany

Background.  Infections by multidrug-resistant Pseudomonas aeruginosa (MDRPa) are an important cause of morbidity and
mortality in patients after allogeneic hematopoietic stem cell transplantation (HSCT). Humid environments can serve as a reservoir
and source of infection by this pathogen. To minimize the risk of infection from these reservoirs, we performed extensive remodeling
of sanitation and water installations as the focus of our hygiene bundle.

Methods.  During the reconstruction of our transplantation unit (April 2011–April 2014)  we implemented several technical
modifications to reduce environmental contamination by and subsequent spreading of MDRPa, including a newly designed shower
drain, disinfecting siphons underneath the sinks, and rimless toilets. During a 3-year study period (2012–2014), we tracked the
number of patients affected by MDRPa (colonized and/or infected) and the outcome of infected patients, and monitored the envi-
ronmental occurrence of this pathogen. We further performed whole-genome sequencing of nosocomial MDRPa strains to evaluate
genotypic relationships between isolates.

Results.  Whereas 31 (9.2%; 18 colonized, 13 infected) patients were affected in 2012 and 2013, the number decreased to 3 in
2014 (17%; 3 colonized, 0 infected). Lethality by MDRPa similarly decreased from 3.6% to 0%. Environmental detection of MDRPa
decreased in toilets from 18.9% in 2012–2013 to 6.1% in the following year and from 8.1% to 3.0%, respectively, in shower outlets.
Whole-genome sequencing showed close relationships between environmental and patient-derived isolates.

Conclusions.  Hospital construction measures aimed at controlling environmental contamination by and spread of MDRPa are
effective at minimizing the risk of highly lethal MDRPa infections.

Keywords.  multidrug-resistant Pseudomonas aeruginosa; infection control; hospital acquired infection; immunocompromised
hosts; hospital water system.

Resistance rates in gram-negative bacteria are increasing and In outbreak situations, various environmental reservoirs
pose a threat for patients with hematologic malignancies have been identified. The sanitary and water supply system
[1]. Patients undergoing allogeneic hematopoietic stem cell is a common reservoir for P.  aeruginosa biofilm formation
transplantation (HSCT) are especially at risk for acquiring [9]. As a consequence, hospital construction design needs to
Pseudomonas aeruginosa infections due to their immunocom- consider these sources of transmission of MDRPa from the
promised state [2], which also puts them at higher risk for a environment. We hypothesized that environmental reservoirs
fatal outcome [3, 4]. Multidrug-resistant Pseudomonas aerug- pose a severe risk to patients undergoing HSCT for acquiring
inosa (MDRPa) outbreak situations were repeatedly described MDRPa and developing invasive disease. Therefore, during a
on hematological-oncological wards [5–7] and were associated substantial reconstruction and expansion of our transplanta-
with a death rate of up to 80% [8]. tion unit starting in 2011, we implemented a bundle of meas-
ures that included a variety of hygiene measures with the
Received 5 January 2017; editorial decision 3 May 2017; accepted 15 May 2017; published major focus on comprehensive renovation and a new design
online May 18, 2017. of the sanitary and water supply system to ultimately reduce
the number of MDRPa infections. Over the following 3 years,
aA. M. and M. S. contributed equally to this work. in which the measures in hospital construction design were
Correspondence: M.  Stelljes, Department of Medicine A, Hematology and Oncology, performed, we studied microbiological surveillance data,
University of Muenster, Albert-Schweitzer-Campus 1, 48149 Muenster, Germany (stelljes@ determined the relationship between MDRPa isolated from
uni-muenster.de). both patients and the environment using whole-genome

Clinical Infectious Diseases®  2017;65(6):935–42
© The Author 2017. Published by Oxford University Press for the Infectious Diseases Society
of America. All rights reserved. For permissions, e-mail: [email protected].
DOI: 10.1093/cid/cix465

MDR P. aeruginosa Infection Control  •  CID 2017:65 (15 September) • 935

sequencing (WGS), and monitored patient outcome with renovation of the existing ward was started at the same time
respect to MDRPa infections. and completed in April 2014.

METHODS During the observation period of this study from January
2012 to December 2014, 474 pediatric and adult patients under-
Patients and Treatment went HSCT on either ward of the unit, accounting for a total of
517 admissions, including 43 readmissions for transplant-asso-
The Center for Bone Marrow Transplantation of the ciated complications. Key demographic and medical character-
University Hospital Muenster serves a catchment area of 5 istics of the 474 patients are summarized in Table 1. In general,
million people in the northwest of Germany. It consists of patients undergoing allogeneic HSCT are treated for 4–6 weeks
an outpatient clinic and an interdisciplinary inpatient unit as inpatients. Standard antibacterial prophylaxis consisted of
with 10 high-efficiency particulate air (HEPA)–filtered sin- ciprofloxacin starting upon admission; the standard regimen
gle-patient rooms, all of which were equipped with separate for empiric antibacterial therapy for neutropenic fever consisted
bathrooms and an anteroom. In March 2011, the inpatient of piperacillin-tazobactam plus gentamicin. Death caused by
unit was extended by a second, similarly constructed ward an MDRPa infection was diagnosed by the treating physicians
resulting in a total capacity of 20 rooms. Room-by-room

Table 1.  Characteristics of Patients Receiving an Allogeneic Hematopoietic Stem Cell Transplant in the Transplantation Unit, 2012–2014 Downloaded from https://academic.oup.com/cid/article/65/6/935/3831257 by guest on 09 May 2022

Characteristics of HSCT Patients Patients in 2012 (n = 161) Patients in 2013 (n = 153) Patients in 2014 (n = 158)
135/23 (85/15)
Adult/pediatric patients, no./No. (%) 133/28 (83/17) 132/21 (86 /14)
56 (20–73)
Median age at HSCT, y (range) 54 (20–77) 16 (0.7–19)
13 (0.7–19) 99 (63)
  Adult patients 53 (19–71) 106 (69)
63
  Pediatric patients 10 (0.7–20) 66 19
26 22
Male sex, No. (%) 106 (66) 17 15/2
13/6 8
Diagnosis, No. 8 4
5 6
  Acute myeloid leukemia 82 3 9
3 7
  Acute lymphatic leukemia 28 4 3
2
  Myelodysplastic syndrome 9 27 (17)
37 (24) 16 (10)
  Non-Hodgkin/Hodgkin lymphoma 11/0 21 (14) 86 (54)
74 (48) 29 (18)
  Multiple myeloma 8 21 (14)
40 (25)
  Chronic lymphatic leukemia 7 50 (33)
1 118 (75)
  Chronic myeloid leukemia 6 103 (67)
51 (32)
 Myelofibrosis 2 47 (31) 71 (45)
71 (46) 36 (23)
  Aplastic anemia 1 32 (21)
3 7 (4–12)
 Other 7 15 (11–22)
9 (5–14)
Remission status prior to HSCT, No. (%) 22 (17–30) 82 (76–88)
67 (59–74)
  First complete remission 50 (34) 77 (70–83) 62 (54–69)
58 (50–66)
  2nd or 3rd complete remission 17 (11) 53 (45–61)

  Active/refractory disease 79 (49)

  Primary disease, untreateda 15 (9)

Donor, No. (%)

  Related (sibling) 44 (27)

  Haploidentical 1

 Unrelated 117 (73)

Conditioning regimens, No. (%)

  Full-toxic conditioning 60 (37)

  Dose-adapted/reduced intensity conditioning 40 (25)

  Sequential conditioning (refractory/active AML) 59 (37)

 Other 2

Cumulative incidence of nonrelapse death rate after HSCT, % (95% CI)

  At day +30 3 (1–7)

  At day +100 9 (6–15)

Overall survival, % (95% CI)

  At day +100 87 (83–93)

  At 1 y 63 (56–71)

  At 2 y 57 (49–64)

Abbreviations: AML, acute myeloid leukemia; CI, confidence interval; HSCT, hematopoietic stem cell transplantation.
aFor example, myelodysplastic syndrome without specific treatment before HSCT.

936 • CID 2017:65 (15 September) •  Kossow et al

in the case of detection of MDRPa in blood cultures directly a system that provides the flushing water with disinfectant con-
obtained from patients with a severe sepsis, which resulted taining 0.5% glucoprotamin.
in death within a few hours (no evidence for other causes/no
detection of other bacteria). Disinfection and Sterilization

The causality of death after allogeneic HSCT was defined Patient rooms, including all bathroom items, were cleaned and
as being non–relapse related in cases of an ongoing remission disinfected on a daily basis and when a patient was discharged
after transplantation and relapse related in the case of a relapsed during the entire study period. In addition, surfaces with frequent
disease after transplantation, even in cases with a subsequent hand contact were disinfected repeatedly. The bubble cap and
remission. The exact cause of death, especially of patients with heavy lid of the shower drains were sterilized weekly and after the
a longer follow-up after transplantation, is usually only known discharge of every patient. Disinfectants contained either quater-
in a small proportion of cases, as the treatment of late complica- nary ammonium compounds or glucoprotamin at a concentra-
tion was also performed outside the transplant center. tion of 0.5%. These practices continue to be implemented.

Hospital Construction Sampling Downloaded from https://academic.oup.com/cid/article/65/6/935/3831257 by guest on 09 May 2022

To reduce the risk of biofilm formation and subsequent trans- Patients were screened for MDRPa upon admission and weekly
mission of bacterial pathogens to the patients, we developed a thereafter during their stay via rectal swabs starting in April 2011
new shower drain design. These drains are characterized by a as part of our routine surveillance strategy for the control of mul-
15.3 cm wide and 36 cm deep outlet, which is easy to clean and tidrug-resistant (MDR) bacteria. Colonization and infection were
disinfect (Figure 1). To prevent aerosol formation, the drain is considered to be nosocomial when screening upon admission was
covered by a heavy stainless steel lid to discourage intentional negative and/or MDRPa was isolated >48 hours after admission.
and prevent accidental removal by patients (Figure 1A). Part of
the design includes a stainless steel bubble cap, which prevents In parallel, we began a systematic environmental sampling
the spread of odors (Figure 1B). This bubble cap can be removed including every new room upon completed renovation. The
for cleaning, disinfection, and sterilization (Figure 1C). environmental sampling included swabs from shower out-
lets and washbasins as well as toilets starting in March 2012.
Shower heads and faucets were equipped with Pall Aquasafe Sampling took place at least once a month in every room and
water filters (Pall Cooperation, Port Washington, New York). following the discharge of a colonized or infected patient.
We installed Biorec disinfection siphons (Biorec Dr. Schluttig,
Lauta, Germany) under all washbasins (Figure 2A). These Identification of Bacteria
siphons use ultraviolet light, frequent vibration (50–200 Hz),
temperature (85°C), and an antibacterial coating to prevent Samples were plated on Columbia sheep blood agar and
biofilm formation. Pseudomonas selective cetrimide agar (both Oxoid, Wesel,
Germany). Environmental swabs were enriched in tryptic soy
Because we had also detected MDRPa in toilets, we also broth for 24 hours at 36°C before being plated again on Columbia
decided to replace all toilets with rimless toilets (Figure 2B) and sheep blood agar. Species identification was confirmed by

Figure 1.  Shower drain design. The whole installation is covered by a heavy stainless steel lid (A), which is designed to discourage patients from opening and to prevent
accidental removal. The drain (B) is designed to be wide, so it can be easily cleaned and disinfected. A bubble cap insert (C) prevents odors and splashing and can be removed
for sterilization.

MDR P. aeruginosa Infection Control  •  CID 2017:65 (15 September) • 937

Figure 2.  Hygiene siphon and rimless toilet. The Hygiene siphon uses vibration (50–200 Hz), high temperatures (85°C), and ultraviolet light to prevent biofilm formation. It Downloaded from https://academic.oup.com/cid/article/65/6/935/3831257 by guest on 09 May 2022
is covered to discourage a frequent change of settings and for cleaning and disinfection purposes.

matrix-assisted laser desorption/ionization (MALDI) time-of- Infection Control Measures
flight mass spectrometry using the MALDI-Biotyper system
(Bruker Daltonics, Bremen, Germany). We performed the anti- According to the national guideline [10], every patient col-
biotic susceptibility testing using VITEK 2 (bioMérieux, Marcy- onized or infected with any MDR bacteria undergoes con-
l’Étoile, France) for patient-derived isolates and disk diffusion tact isolation. An infection control nurse was present on the
for environmental isolates; both were interpreted in accordance ward at least once a week and ward rounds were accompanied
to the European Committee on Antimicrobial Susceptibility by an infection control physician. Moreover, a pharmacist
Testing breakpoints (version 2.0, 29 October 2011). Isolates is present on the ward 3 times a week. Since January 2013
resistant against piperacillin, third-/fourth-generation ceph- a supervisor who underwent special training and stood in
alosporins, fluoroquinolones, and carbapenems were rated as close contact with the infection control team was added to
MDR in accordance with the recommendations set forth by the the cleaning team.
national German Robert Koch Institute [10].
Statistical Analysis
Whole-Genome Sequencing–Based Typing
All patient and environmental data are expressed as absolute
Whole-genome sequencing was used to determine possible numbers or percentages, if not stated otherwise. Statistical anal-
clonal relationships between environmental and patient-de- yses were performed using the Fisher’s exact test for categorical
rived isolates from the transplantation unit as part of our routine data. Statistical significance was considered using a P value of
surveillance strategy. DNA extraction, WGS library prepara- <.05. For the analysis of the effectiveness of the implemented
tion, sequencing, and subsequent data analysis were performed measures on colonization and infection rates, we compared
as recently described [11]. Using P.  aeruginosa strain PAO1 the occurrence in patients and the environment in the years
(GenBank accession number NC_002516) as the reference 2012–2013 with our findings in 2014. To estimate accuracy, the
sequence, coding regions were compared in a gene-by-gene relative risk and 95% confidence interval were calculated for
approach based on up to 3842 genes [11] using the SeqSphere+ affected patients and the lethality by MDRPa.
software version 2 (Ridom GmbH, Muenster, Germany).
Isolates that differed in a pairwise comparison in >14 alleles RESULTS
were assumed to be unrelated [11]. The clonal relationship is
displayed in a minimum-spanning tree. For backwards com- Occurrence in Patients
patibility with classical molecular typing—that is, multilocus
sequence typing (MLST)—the MLST sequence types (STs) were One hundred seventy-one and 167 patients were screened
extracted from the WGS data in silico. for MDRPa in 2012 and 2013, respectively. They were com-
pared to the 179 patients treated in 2014. The number of
patients nosocomially affected (colonized and/or infected)
by MDRPa decreased from 31 (9.17%) in 2012–2013 to 3

938 • CID 2017:65 (15 September) •  Kossow et al

Figure 3.  Comparison of occurrences of multidrug-resistant Pseudomonas aeruginosa (MDRPa) in the environment and patients. Bars of different shades give the percent- Downloaded from https://academic.oup.com/cid/article/65/6/935/3831257 by guest on 09 May 2022
age of occurrence of MDRPa in toilets, shower outlets, and patients. Absolute numbers are given on top of each column for 2012–2013 and 2014. The pairwise comparisons

and statistical analysis between the findings of the years 2012–2013 and 2014, when all renovation was completed, are shown in vertical lines. *P < .01; **P < .001.

(1.68%) in 2014 (P < .001; Figure 3). In total, 13 patients were Environmental Detection
infected in 2012 and 2013 (3.85%); this number decreased
to 1 patient in 2014 (0.56%) (P  <  .05). In 2012 and 2013, Environmental sampling detected MDRPa in 125 of 660 sam-
12 patients died of MDRPa infection (3.55%) whereas this ples taken from toilets (18.94%) and 52 of 641 samples taken
number decreased to 0 in 2014 (0%) (P < .05). In the 2 years from shower outlets (8.11%) in the years 2012 and 2013. These
following the initial study period, 8 of 376 patients were numbers decreased to 23 of 375 toilets testing positive (6.13%;
affected, (2.13%) 4 of whom died. The annual distribution P < .001) and 11 of 372 shower drains testing positive (2.96%;
of affected patients and further clinical details (infection and P < .01) in 2014 (Figure 2).
colonization) are shown in Table 2.
During follow-up, we found occurrence of MDRPa in 29 of
725 toilets (4%) as well as 33 of 719 shower drains (5.59%).

Table 2.  Annual Distribution and Clinical Details of Patients Colonized or Infected With Multidrug-Resistant Pseudomonas aeruginosa

No. of Patients With:

Year No. of Patients Treated BSI Pneumonia Colonization Total No. of Affected Patients (%)a Lethality, %b

2012 171 40 14 18 (10.5) 2.3 (n = 4)
2013 167 81 4 13 (7.8) 4.8 (n = 8)
2014 179 10 2
2015 183 40 3 3 (1.7) 0 (n = 0)
2016 193 00 1 7 (3.8) 2.2 (n = 4)
1 (0.5)
0 (n = 0)

Abbreviation: BSI, bloodstream infection.
aP < .001 (relative risk [RR], 5.5; 95% confidence interval [CI], 1.6–22.3).
bP < .05 (RR, 132.4; 95% CI, 1.3–8507.7).

MDR P. aeruginosa Infection Control  •  CID 2017:65 (15 September) • 939

Sinks tested positively in 6 of 641 (0.93%) samples in 2012– relationship and differed from those derived from patients
2013 and not at all in 2014. Occurrence in sinks could always be treated on other wards of our hospital. Isolates from patients
directly linked to a defected disinfection siphon. who were treated on oncological wards other than the HSCT
ward were either closely or not at all related to the cluster
Whole-Genome Sequencing from the HSCT ward. Extraction of MLST STs from the WGS
data resulted in 51 different STs; of these, the most common
In total, 187 MDRPa isolates were analyzed by WGS ST in patients was ST235 (n = 35) (Supplementary Table 1).
(Supplementary Table  1). In this analysis we included 151 Comparison of WGS-based allelic profiles of up to 3842 genes
patients’ and 36 randomly chosen environmental samples. (Figure  4) showed a close relationship between patient-de-
All environmental isolates stemmed from our HSCT ward. Of rived isolates and environmental isolates in the majority of
the 151 patient-derived isolates, 19 originated from patients cases. In contrast, isolates from wards other than the trans-
who were treated on oncological wards other than the HSCT plantation unit were mostly unrelated, suggesting a relation-
unit and 32 from HSCT patients. The remaining 100 sam- ship between patient-derived and environmental isolates on
ples came from nononcological wards of the same hospi- the transplantation unit.
tal. Isolates from the HSCT ward mostly showed a genetic

Downloaded from https://academic.oup.com/cid/article/65/6/935/3831257 by guest on 09 May 2022

Figure 4.  Minimum spanning tree of 187 MDRPa isolates either from patients (P) or environmental samples (E) originating from our transplantation unit and other onco-
logical or nononcological wards of our hospital. Each circle represents a single genotype, that is, an allelic profile based on up to 3842 target genes present in the isolates
with the “pairwise ignoring missing values” option turned on in the SeqSphere+ software during comparison. The circles are named with the isolate ID(s) and colored by ward
and origin. Thick connecting lines between 2 genotypes display ≤14 differing alleles; thin lines display genotypic distances of >14 alleles’ distance. Abbreviations: AHSCT,
allogeneic hematopoietic stem cell transplantation.

940 • CID 2017:65 (15 September) •  Kossow et al

DISCUSSION Previously, pulsed-field gel electrophoresis was frequently Downloaded from https://academic.oup.com/cid/article/65/6/935/3831257 by guest on 09 May 2022
used for the detection of genetic relationships between strains
Pseudomonas aeruginosa infection is a severe complication in the isolated from different patients in suspected outbreak situa-
course of allogeneic HSCT. Different strategies were found to be tions [5, 8, 12, 13, 17]; however, this methodology is nowadays
effective in identifying and eliminating sources of this pathogen, increasingly replaced by WGS-based typing, which provides
thus preventing further transmissions. These include hygiene an even higher discriminatory genotyping [14, 19–22]. The
measures such as contact precautions [12], strengthening hand introduction of WGS during the study period aimed to reveal
hygiene practices [13, 14], and the screening of patients [7, 14] an answer to this open question as WGS provides much more
as well as the installation of point-of-use water filters [7, 15, 16]. precise genomic comparison. As a result, a contamination from
the environment could not be conclusively excluded.
After detection of MDRPa in patients with fatal bloodstream
infections, we did an extensive surveillance of common reservoir The case fatality rate among infected patients was high, and
sites. As several isolates were detected in the patients’ environment it became obvious during the course of our study that coloniza-
and, in particular, the water supply and sanitary system, acqui- tion by MDRPa is often detected only shortly before the death
sition of the organism from the environment was considered as of a patient. This is confirmed by other reports of highly viru-
the probable source of colonization and infection. Consequently, lent MDRPa clones, including MLST STs 111, 175, and 235 [23].
we applied extensive structural measures in combination with MLST can therefore provide risk analysis for the virulence of sin-
intensified standard cleaning procedures and intensive repetitive gle isolates. Beyond this, WGS grants the possibility to reconstruct
training of all staff with direct or indirect patient contact, includ- genetic relationships in outbreak situations in order to concen-
ing the cleaning and housekeeping personnel. Those measures trate hygiene measures upon the original source. Of note, we also
were bundled together with prospective patient screening and found a high number of ST235 in our isolates (Supplementary
the regular presence of an infection control nurse. This bundle Table 1). The rapid evolution from colonization to severe infec-
led to a decrease of MDRPa detections in both the environment tion is worrisome given that other groups found infection to be
(shower outlets and toilets) and patients. Occurrence in wash- more likely in patients who were previously colonized [24].
basins was low in both periods and always directly linked to a
nonfunctioning disinfection siphon. Disinfection siphons are an The approach of implementing a bundle of infection con-
important part of our bundle, but their reliability was not ques-
tioned during the course of our study. trol measures as the most effective strategy for the prevention

Most important, lethality due to MDRPa infections of infection by MDR bacteria in patients with HSCT is also
decreased to 0% at the end of the study period. By elucidat-
ing the close relationship between almost all MDRPa isolates supported by the review of Ruhnke et  al [4]. Given the envi-
from the transplantation unit, WGS underlined the impact of
the interplay between the patient and the environment on the ronmental presence of MDRPa, we added environmental
one hand and the importance to reduce or even eliminate this
reservoir on the other hand. In contrast to other settings [5, 7, decontamination as the preventive measure for MDRPa infec-
9, 13, 15–18], we were not able to determine a specific single
source for colonization and infection. There was no obvious tions. Antibiotic prophylaxis and first-line treatment did not
correlation with specific patient rooms or the time of occur-
rence between environmental isolates and/or patient-derived change during the study period. Hand hygiene compliance was
isolates (Supplementary Figure  1). Other observations [6, 8]
also were unable to differentiate whether an individual patient observed to be high by the infection control nurse and hand
was colonized by the environment or the environment was con-
taminated by an already colonized patient. hygiene training was provided continuously during the entire

Even after screening all patients for MDRPa at the time of study period. Given the obvious success of these measures, the
admission to the transplant unit for MDRPa, it remains unclear
to which extent patients acquired MDRPa during their hospital study provides suggestions for specific preciously undescribed
stay or were colonized before. Probably, standard microbiologi-
cal screening methods might be not sensitive enough to detect infection control considerations for the construction of patients’
MDRPa in patients with normal gut colonization. In contrast to
the environmental screening swabs, we did not use broth enrich- bathrooms, with a reciprocal effect to classic infection control
ment on screening swabs from patients even though selective
agar plates were used in both cases. Using broth enrichment measures. Even though the individual parts of our bundle are
might lead to an earlier detection of MDRPa colonization.
meant to enhance each other, we attribute the main effects to

constructional changes that were specifically aiming at prevent-

ing MDRPa. A retrospective passive surveillance of bloodstream

infections caused by other resistant pathogens shows a low

baseline and no decline in infection rates (methicillin-resistant

Staphylococcus aureus: n = 0 in 2012–2013, n = 2 in 2014; vanco-

mycin-resistant enterococci: n = 2 in 2012–2013, n = 5 in 2014).
The study is limited because of the fact that no systematic

data exist concerning environmental occurrence and patients
affected prior to the start of the remodeling of the water supply
and sanitary system. Moreover, in some cases, patients could
not been screened weekly due to their clinical status. This might
have led to a missed MDRPa colonization before the actual

MDR P. aeruginosa Infection Control  •  CID 2017:65 (15 September) • 941

detection. A detailed analysis of epidemiological and clinical 3. Patel SJ, Oliveira AP, Zhou JJ, et al. Risk factors and outcomes of infections caused Downloaded from https://academic.oup.com/cid/article/65/6/935/3831257 by guest on 09 May 2022
data might help to identify predictive parameters for patients by extremely drug-resistant gram-negative bacilli in patients hospitalized in
to acquire colonization or prediction for the clinical outcome of intensive care units. Am J Infect Control 2014; 42:626–31.
infection. Considering the low number of affected patients and
the heterogeneous course of underlying disease and treatment, 4. Ruhnke M, Arnold R, Gastmeier P. Infection control issues in patients with hae-
this would not lead to reliable results in our setting. matological malignancies in the era of multidrug-resistant bacteria. Lancet Oncol
2014; 15:e606–19.
It is well known that in other cohorts, patients exchange iso-
lates among each other. It may be evaluated whether this also 5. Engelhart S, Krizek L, Glasmacher A, Fischnaller E, Marklein G, Exner M.
happens in outpatient settings as other settings than the inpa- Pseudomonas aeruginosa outbreak in a haematology-oncology unit associated
tient clinic are considered to possibly contribute to transmission with contaminated surface cleaning equipment. J Hosp Infect 2002; 52:93–8.
[4, 25]. In this respect, we also consider reviewing our screening
methods to improve selectivity and achieve an earlier detection, 6. Gillespie TA, Johnson PR, Notman AW, Coia JE, Hanson MF. Eradication of a
that is, while still in the state of colonization. In the 2 years fol- resistant Pseudomonas aeruginosa strain after a cluster of infections in a hematol-
lowing the study period, we experienced a higher number of ogy/oncology unit. Clin Microbiol Infect 2000; 6:125–30.
preexisting colonizations in patients admitted to our transplan-
tation unit (n = 3 each in 2015 and 2016). This increased colo- 7. Vianelli N, Giannini MB, Quarti C, et al. Resolution of a Pseudomonas aerugi-
nization pressure might have been a reason for the slight rise in nosa outbreak in a hematology unit with the use of disposable sterile water filters.
occurrence in patients (3.82%) and toilets (7.61%) in 2015. As a Haematologica 2006; 91:983–5.
reaction to this, we performed extensive maintenance of toilets.
In 2016, we could not detect MDRPa in toilets, and nosocomial 8. Mudau M, Jacobson R, Minenza N, et  al. Outbreak of multi-drug resistant
occurrence in patients decreased to 0.5%. Overall occurrence Pseudomonas aeruginosa bloodstream infection in the haematology unit of a
remained low for a prolonged period of time. South African academic hospital. PLoS One 2013; 8:e55985.

Taken together, our study demonstrates that structural 9. Loveday HP, Wilson JA, Kerr K, Pitchers R, Walker JT, Browne J. Association
changes combined with a bundle of infection prevention meas- between healthcare water systems and Pseudomonas aeruginosa infections: a
ures can lead to a decrease in environmental occurrence and a rapid systematic review. J Hosp Infect 2014; 86:7–15.
reduction of MDRPa infections in HSCT patients. This strategy,
with its special focus on hospital construction measures, may 10. Robert Koch Institute. Hygienemaßnahmen bei Infektionen oder Besiedlung mit
serve as a model for other institutions facing the daunting chal- multiresistenten gramnegativen Stäbchen [in German]. Bundesgesundheitsblatt
lenge of MDR bacteria in high-risk patients. 2012; 55:1311–54.

Supplementary Data 11. Mellmann A, Bletz S, Böking T, et  al. Real-time genome sequencing of resist-
ant bacteria provides precision infection control in an institutional setting. J Clin
Supplementary materials are available at Clinical Infectious Diseases online. Microbiol 2016; 54:2874–81.
Consisting of data provided by the authors to benefit the reader, the posted
materials are not copyedited and are the sole responsibility of the authors, 12. Nagao M, Iinuma Y, Igawa J, et al. Control of an outbreak of carbapenem-resist-
so questions or comments should be addressed to the corresponding author. ant Pseudomonas aeruginosa in a haemato-oncology unit. J Hosp Infect 2011;
79:49–53.
Notes
13. Blanc DS, Nahimana I, Petignat C, Wenger A, Bille J, Francioli P. Faucets as a res-
Acknowledgments.  We thank Thomas Boeking, Isabell Höfig, Margret ervoir of endemic Pseudomonas aeruginosa colonization/infections in intensive
Junge, and Sven Rottmann for skillful technical assistance; Sabine Hahn, care units. Intensive Care Med 2004; 30:1964–8.
Roswitha Wiersbin, and Brigitte Winckler for environmental sampling;
Bernhard Fiedler for the clinical data collection; and Eric J. Bernhard for 14. Willmann M, Bezdan D, Zapata L, et  al. Analysis of a long-term outbreak of
proofreading the manuscript. XDR Pseudomonas aeruginosa: a molecular epidemiological study. J Antimicrob
Chemother 2015; 70:1322–30.
Potential conflicts of interest.  All authors: No reported conflicts of
interest. All authors have submitted the ICMJE Form for Disclosure of 15. Trautmann M, Halder S, Hoegel J, Royer H, Haller M. Point-of-use water filtra-
Potential Conflicts of Interest. Conflicts that the editors consider relevant to tion reduces endemic Pseudomonas aeruginosa infections on a surgical intensive
the content of the manuscript have been disclosed. care unit. Am J Infect Control 2008; 36:421–9.

References 16. Aumeran C, Paillard C, Robin F, et al. Pseudomonas aeruginosa and Pseudomonas
putida outbreak associated with contaminated water outlets in an oncohaematol-
1. Blennow O, Ljungman P. The challenge of antibiotic resistance in haematology ogy paediatric unit. J Hosp Infect 2007; 65:47–53.
patients. Br J Haematol 2016; 172:497–511.
17. Hota S, Hirji Z, Stockton K, et al. Outbreak of multidrug-resistant Pseudomonas
2. Kerr KG, Snelling AM. Pseudomonas aeruginosa: a formidable and ever-present aeruginosa colonization and infection secondary to imperfect intensive care unit
adversary. J Hosp Infect 2009; 73:338–44. room design. Infect Control Hosp Epidemiol 2009; 30:25–33.

18. Quick J, Cumley N, Wearn CM, et al. Seeking the source of Pseudomonas aerug-
inosa infections in a recently opened hospital: an observational study using
whole-genome sequencing. BMJ Open 2014; 4:e006278.

19. Snyder LA, Loman NJ, Faraj LA, et  al. Epidemiological investigation of
Pseudomonas aeruginosa isolates from a six-year-long hospital outbreak using
high-throughput whole genome sequencing. Euro Surveill 18.

20. Köser CU, Ellington MJ, Cartwright EJ, et  al. Routine use of microbial whole
genome sequencing in diagnostic and public health microbiology. PLoS Pathog
2012; 8:e1002824.

21. Köser CU, Holden MT, Ellington MJ, et al. Rapid whole-genome sequencing for
investigation of a neonatal MRSA outbreak. N Engl J Med 2012; 366:2267–75.

22. Harris SR, Cartwright EJ, Török ME, et al. Whole-genome sequencing for analysis
of an outbreak of meticillin-resistant Staphylococcus aureus: a descriptive study.
Lancet Infect Dis 2013; 13:130–6.

23. Oliver A, Mulet X, López-Causapé C, Juan C. The increasing threat of
Pseudomonas aeruginosa high-risk clones. Drug Resist Updat 2015; 21–22:41–59.

24. Willmann M, Klimek AM, Vogel W, et  al. Clinical and treatment-related risk
factors for nosocomial colonisation with extensively drug-resistant Pseudomonas
aeruginosa in a haematological patient population: a matched case control study.
BMC Infect Dis 2014; 14:650.

25. Dobbs TE, Guh AY, Oakes P, et  al. Outbreak of Pseudomonas aeruginosa and
Klebsiella pneumoniae bloodstream infections at an outpatient chemotherapy
center. Am J Infect Control 2014; 42:731–4.

942 • CID 2017:65 (15 September) •  Kossow et al

PUBLIC AND ENVIRONMENTAL
HEALTH MICROBIOLOGY

crossm

Droplet- Rather than Aerosol-Mediated Dispersion Is the
Primary Mechanism of Bacterial Transmission from
Contaminated Hand-Washing Sink Traps

Shireen M. Kotay,a Rodney M. Donlan,b Christine Ganim,b Katie Barry,a Bryan E. Christensen,b Amy J. Mathersa,c

aDivision of Infectious Diseases and International Health, Department of Medicine, University of Virginia Health System, Charlottesville, Virginia, USA
bDivision of Healthcare Quality Promotion, Centers for Disease Control and Prevention, Atlanta, Georgia, USA
cClinical Microbiology, Department of Pathology, University of Virginia Health System, Charlottesville, Virginia, USA

ABSTRACT An alarming rise in hospital outbreaks implicating hand-washing sinks Citation Kotay SM, Donlan RM, Ganim C, Barry Downloaded from https://journals.asm.org/journal/aem on 26 March 2022 by 2a01:6500:a036:1f2b:3aae:6191:1bf:901.
has led to widespread acknowledgment that sinks are a major reservoir of antibiotic- K, Christensen BE, Mathers AJ. 2019. Droplet-
resistant pathogens in patient care areas. An earlier study using green fluorescent rather than aerosol-mediated dispersion is the
protein (GFP)-expressing Escherichia coli (GFP-E. coli) as a model organism demon- primary mechanism of bacterial transmission
strated dispersal from drain biofilms in contaminated sinks. The present study fur- from contaminated hand-washing sink traps.
ther characterizes the dispersal of microorganisms from contaminated sinks. Repli- Appl Environ Microbiol 85:e01997-18. https://
cate hand-washing sinks were inoculated with GFP-E. coli, and dispersion was doi.org/10.1128/AEM.01997-18.
measured using qualitative (settle plates) and quantitative (air sampling) methods. Editor Andrew J. McBain, University of
Dispersal caused by faucet water was captured with settle plates and air sampling Manchester
methods when bacteria were present on the drain. In contrast, no dispersal was cap- This is a work of the U.S. Government and is
tured without or in between faucet events, amending an earlier theory that bacteria not subject to copyright protection in the
aerosolize from the P-trap and disperse. Numbers of dispersed GFP-E. coli cells di- United States. Foreign copyrights may apply.
minished substantially within 30 minutes after faucet usage, suggesting that the or- Address correspondence to Amy J. Mathers,
ganisms were associated with larger droplet-sized particles that are not suspended [email protected].
in the air for long periods. Received 14 August 2018
Accepted 18 October 2018
IMPORTANCE Among the possible environmental reservoirs in a patient care en- Accepted manuscript posted online 26
vironment, sink drains are increasingly recognized as a potential reservoir to hos- October 2018
pitalized patients of multidrug-resistant health care-associated pathogens. With Published 9 January 2019
increasing antimicrobial resistance limiting therapeutic options for patients, a
better understanding of how pathogens disseminate from sink drains is urgently aem.asm.org 1
needed. Once this knowledge gap has decreased, interventions can be engi-
neered to decrease or eliminate transmission from hospital sink drains to pa-
tients. The current study further defines the mechanisms of transmission for bac-
teria that colonize sink drains.

KEYWORDS antimicrobial resistance, dispersion, GFP-E. coli, sink lab, carbapenem-
resistant Enterobacteriaceae, hand washing, infection control, nosocomial
transmission, premise plumbing, sinks

Recent reports have implicated hand-washing sinks as a primary reservoir of
antibiotic-resistant pathogens within patient care environments (1–27). Many of
these reports have been published since 2016, highlighting the global recognition and
vital role that biofilms located in and on sinks can have in disseminating clinically
important, drug-resistant, Gram-negative bacteria (14–27). Retrospective and prospec-
tive surveillance investigations affirm that hospital sinks provide habitats for several
opportunistic pathogens, raising serious concerns (8, 24, 26, 28). It is not the mere
presence of these drug-resistant pathogens in the hospital wastewater that is of
concern, but the ability of these organisms to colonize biofilms on the luminal

January 2019 Volume 85 Issue 2 e01997-18 Applied and Environmental Microbiology

Kotay et al. Applied and Environmental Microbiology
aem.asm.org 2
TABLE 1 Chlorine concentrations and faucet water temperaturea

Parameter Initial (first catch) Final (after 2 min)

Total chlorine (mg/liter) 0.77 Ϯ 0.26 0.63 Ϯ 0.07
Free chlorine (mg/liter) 0.62 Ϯ 0.25 0.54 Ϯ 0.13
Water temperature (°C) 22.53 Ϯ 1.87 36.69 Ϯ 2.17

aMeans and standard deviations of total and free chlorine residual concentrations and water temperature,
measured at different time points and from water collected from faucets supplying different sinks over the
course of this study.

surfaces of wastewater plumbing and thereby withstand routine cleaning practices. Downloaded from https://journals.asm.org/journal/aem on 26 March 2022 by 2a01:6500:a036:1f2b:3aae:6191:1bf:901.
While several Gammaproteobacteria species detected from the sinks in hospitals
have been linked to health care-associated infections, opportunistic pathogens like
Pseudomonas aeruginosa, Acinetobacter baumannii, and Stenotrophomonas malto-
philia are typically known to be found in water environments (29–31). In contrast,
emerging pathogens, such as carbapenemase-producing Enterobacteriaceae (CPE),
many of which have fecal origins, may survive within the biofilm formed on sink
surfaces and wastewater premise plumbing (32, 33). CPE infections are especially
threatening because they often acquire mobile resistance elements through hori-
zontal gene transfer and are more frequent causes of highly antibiotic-resistant
infections with reduced treatment options.

In outbreak investigations, species and strain matches between patient and sink
isolates are often attributed to sink source contamination; however, the direction (sink
to patient versus patient to sink) and precise mode of transmission remain inconsistent
and elusive (29, 30). Even with increased recognition of transmission, a knowledge gap
exists with regard to the precise mechanism of transmission from sink reservoirs to the
patient. Using a model of sink colonization with green fluorescent protein (GFP)-
expressing Escherichia coli (GFP-E. coli), we recently demonstrated the source and the
degree of dispersion from sink wastewater to the surrounding environment (34).
Factors effecting the rate and extent of droplet-mediated dispersion were investigated,
but particle size involved in dispersion was not measured in this study. Studies that
claim aerosols as the primary dispersion mechanism of bacteria from sinks are based on
rudimentary findings (2, 23, 35, 36) or on assumptions drawn based on these unsub-
stantiated findings (3, 6, 10, 13, 21, 37). Airborne particles originating from sinks can
vary in size and composition. The World Health Organization and Healthcare Infection
Control Practices Advisory Committee (HICPAC) guidelines use a particle diameter of
5 ␮m to delineate between bioaerosol (Յ5 ␮m) and droplet (Ͼ5 ␮m) transmission
(38, 39).

Aerosol-mediated transmission and droplet-mediated transmission in the health
care environment require conceptually different infection control strategies. Clarity
regarding aerosol- versus droplet-mediated dispersion in the context of sinks is critical.
In the present study, we aim to further define the following outstanding knowledge
gaps: (i) the dispersion mechanism of bacteria (aerosol-sized particles or droplets from
biofilms in handwashing sinks), (ii) factors triggering dispersion from a colonized sink
drain, and (iii) the role of biologically active aerosols spontaneously dispersing from the
drain or P-trap without a triggering event. GFP-E. coli as the surrogate organism for
Enterobacteriaceae was used in this model study to investigate these questions.

RESULTS
Free and total chlorine concentrations in the faucet water were consistent across the

experiments (Table 1), as were the water and air temperatures (Table 2). In comparison,
the relative humidity recorded across the experiments varied, with the highest re-
corded in the case of P-trap inoculation experiments (Table 2).

Dispersion immediately following P-trap inoculation. No GFP-E. coli dispersion
was detected on settle plates immediately following inoculation of P-traps with 1010
CFU GFP-E. coli (Fig. 1a). No dispersion was detected using impaction, impingement, or
filtration air sampling methods (Fig. 2). Furthermore, no GFP-E. coli cells were recorded

January 2019 Volume 85 Issue 2 e01997-18

Characterization of Sink Microbial Particles Applied and Environmental Microbiology

TABLE 2 Air temperature and relative humidity recorded across the experimentsa

Type of experiment Air temperature (°C) Relative humidity (%)

Drain inoculation 19.8 Ϯ 0.2 31.4 Ϯ 4.9
P-trap inoculation 19.5 Ϯ 0.4 54.4 Ϯ 11.5
Drain ϩ P-trap colonization 19.9 Ϯ 0.2 44.7 Ϯ 20.3
Control 19.7 Ϯ 0.2 46.5 Ϯ 16.3

aShown are means and standard deviations.

in the P-trap water, P-trap, tailpiece, or sink bowl surface samples collected at the end
of the dispersion experiment.

Dispersion immediately following drain inoculation. When sink drains were
inoculated with 1010 CFU GFP-E. coli, dispersion was detected on settle plates and by
impaction and filtration. Dispersion detected by settle plates across the three sinks
ranged from 35 to 107 CFU/plate. GFP-E. coli levels were higher on the counter space
surrounding the sink bowl than near the faucets (Fig. 1b). Dispersion was not detected
on side-splatter shields. With the exception of the detection of 1 CFU/m3 at 60 min,
using the impaction method, dispersion of GFP-E. coli was detected only at the first
faucet event (t ϭ 0 min), using both impaction and filtration methods (Fig. 2). Average
dispersions captured at the first faucet event (t ϭ 0 min) were 77 and 83 CFU/m3 using

FIG 1 Heat map representation of GFP-E. coli dispersion captured on tryptic soy agar (TSA) settle plates aem.asm.org 3 Downloaded from https://journals.asm.org/journal/aem on 26 March 2022 by 2a01:6500:a036:1f2b:3aae:6191:1bf:901.
following (a) P-trap, (b) sink drain inoculation, and (c) drain line colonization.

January 2019 Volume 85 Issue 2 e01997-18

Kotay et al. Applied and Environmental Microbiology

FIG 2 GFP-E. coli as measured by impaction (SAS90) and gel filtration (MD8) across drain inoculation (D), drain line colonization (D7), and Downloaded from https://journals.asm.org/journal/aem on 26 March 2022 by 2a01:6500:a036:1f2b:3aae:6191:1bf:901.
P-trap inoculation (P) methods.

impaction and filtration methods, respectively. GFP-E. coli was not detected at any time aem.asm.org 4
point using liquid impingement.

Dispersion following growth for 7 days in an amended P-trap biofilm. We
allowed colonization of GFP-E. coli in drain lines between the strainer and P-trap with
nutrient exposure over time, and dispersion was detected on settle plates (Fig. 1c), with
counts ranging from 49 to 107 CFU/plate. The counter space surrounding the sink bowl
received the largest amount of droplet dispersion, followed by faucet, faucet handle
surfaces, and splatter shields. GFP-E. coli levels were highest at the first faucet event
(t ϭ 0 min) and not detectable afterwards, with the exception of a 2 CFU/m3 count at
60 min using air impaction and a 1 CFU/m3 count at 90 min using filtration, similar to
what was observed for the drain inoculation experiment (Fig. 2). Dispersion captured at
the first faucet event (t ϭ 0 min) was 138 and 29 CFU/m3 using impaction and filtration
methods, respectively. GFP-E. coli was not detected using liquid impingement. GFP-E.
coli was detected on the sink bowl, drain grate, tailpiece, and P-trap and in P-trap water
at the completion of this experiment (data not shown).

Dispersion without faucet events (control experiments). Without a faucet event,
GFP-E. coli was not detected on settle plates or by impaction, impingement, or filtration
air sampling methods in any experiment. Without a faucet event, fungal and nonfluo-
rescent CFU were occasionally recorded on settle plates, subsequently identified as
Staphylococcus spp., nonhemolytic streptococci, Paenibacillus spp., yeast, and small
Gram-positive rods.

Total viable heterotrophic organisms in laboratory air. GFP-E. coli was not
detected in the faucet water at any time during the experiments. Air samples were
collected by impaction on Reasoner’s 2A (R2A) medium, with and without faucet
events, in order to quantify total heterotrophic organisms in the air space in proximity
to the sink during each experiment. Across the experiments, the heterotrophic organ-

January 2019 Volume 85 Issue 2 e01997-18

Characterization of Sink Microbial Particles Applied and Environmental Microbiology

FIG 3 Heterotrophic plate counts as measured by impaction air sampling without faucet event (Control) and with faucet events Downloaded from https://journals.asm.org/journal/aem on 26 March 2022 by 2a01:6500:a036:1f2b:3aae:6191:1bf:901.
(Test).

isms in the air ranged from 4 to 578 and 2 to 69 CFU/m3 with (test) and without aem.asm.org 5
(control) faucet events, respectively (Fig. 3). Dispersion captured at the first faucet event
(t ϭ 0 min) when the faucets were turned on (test) was 1 log10 higher than the same
recorded in the case of the control experiment (without faucet event). Heterotrophic
organisms captured from the air steadily declined between the time points 0, 30, 60, 90,
and 120 min, with and without faucet events. Particle concentrations in the air, with
and without faucet events, were found to be consistent during the day of the exper-
iment. However, compared across the experiments, no correlation could be established
(Fig. S1).

Dispersion in the presence of mannequin hands. When mannequin hands, which
created a barrier for water flow onto the drain, were positioned under the faucet and
the sink drain was inoculated with ϳ1010 CFU GFP-E. coli, 10-fold-lower dispersion was
captured on settle plates. No dispersion in air sampling using the impaction or filtration
methods was observed. Dispersion pattern and load on settle plates varied consider-
ably across the three sinks and experiments with the mannequin hands in place.

DISCUSSION
The objective of the present study was to characterize the mechanism of bacterial

dispersion from handwashing sinks, using a GFP plasmid-containing E. coli strain as a
surrogate for multidrug-resistant Enterobacteriaceae. Very few studies have investigated
the dispersion from sinks using methods to sample aerosol-associated microorganisms
(23, 36, 40, 41); however, several studies have drawn subjective interpretations about
aerosol-mediated transmission from contaminated sinks (3, 6, 10, 13, 21, 37). Doring
et al. (41) used an impaction method to examine P. aeruginosa dispersion from the sink
bowl surface during faucet use. P. aeruginosa was detected 15 cm from the sink drain
when counts in the sink drains exceeded 105 CFU/ml (41). Kramer et al. detected 439

January 2019 Volume 85 Issue 2 e01997-18

Kotay et al. Applied and Environmental Microbiology Downloaded from https://journals.asm.org/journal/aem on 26 March 2022 by 2a01:6500:a036:1f2b:3aae:6191:1bf:901.
aem.asm.org 6
CFU/m3 P. aeruginosa in the air sampled 10 cm above the sink drain, with 105 CFU/ml
bacteria in the “sink fluid” (P-trap water) (40). In these studies, dispersion was not
measured without faucet usage (control samples). De Geyter et al. used an MAS-100 air
sampler to measure carbapenem-resistant Enterobacteriaceae (CRE) dispersal from con-
taminated sinks, with and without faucet usage. Several species in the family Entero-
bacteriaceae were detected during faucet usage, but results for control samples (with-
out faucet usage) were not provided (23). Fusch et al., in contrast, sampled air around
sinks with and without faucet events (control) and did not detect P. aeruginosa in air
samples collected without running water (36). We had previously provided a quanti-
tative assessment of dispersal as a function of faucet usage and reported that GFP-E.
coli could be dispersed up to 30 in. beyond the sink drain during faucet usage (34).
However, dispersion was assessed using a gravity method, and bioaerosol production
was not evaluated.

The air sampling methods chosen and tested in the present study were three of the
most widely used methods previously reported (23, 35, 36, 40, 42, 43) and were selected
to assess bioaerosol production during sink usage. In the present study, GFP-E. coli
dispersion was detected during a faucet event but was not detected in the absence of
faucet events using either settle plates or impaction and filtration air sampling meth-
ods. This finding corroborates those of previous studies (2, 21, 34, 36). It also implies
that the shear forces of the faucet water flowing directly onto the sink drain and/or
bowl surfaces results in dispersion of bacteria. Detection of dispersed GFP-E. coli during
a faucet event and nondetection at subsequent time points (after 30 min) suggested
that dispersed cells were associated with larger, heavier droplets that would quickly
settle onto surfaces due to gravity, rather than to aerosol-sized particles that remain in
the air (44). Dispersion of GFP-E. coli from sinks does not appear to be associated with
the production of bioaerosols, that is, of particles smaller than 5 ␮m (35, 40, 45, 46).
Studies that measured air sampling lacked resolution between aerosols and droplets
(35, 40, 41). Air was sampled significantly closer (ϳ4 in.) to the sink drain or impact
point of faucet water on the sink and, therefore, samples might have picked up droplets
rather than aerosols.

A consistent result from this work that is worth reemphasizing is the finding that for
dispersion to occur, the presence of bacteria on the drain and/or bowl surface is
necessary (34). When GFP-E. coli was inoculated into a new P-trap, dispersion was not
detected using settle plates or air sampling methods. This underscores the fact that as
long as the sink drain and bowl remain free of the target organisms (e.g., CPE or other
antibiotic-resistant Gammaproteobacteria), dispersion can be controlled. However, un-
der favorable conditions, bacteria can grow or mobilize from the P-trap into the drain
piping (tailpiece) and colonize the sink drain surfaces, with the potential for a disper-
sion event to occur. This further underlines the importance of sanitary hygiene prac-
tices and strategic surveillance paired with hand-washing-only use of hand-washing
sinks in the patient care environment to reduce the risk of hand-washing sink contam-
ination by multidrug-resistant microorganisms (8). This also emphasizes the necessity
to implement stricter measures to prohibit disposal into the sinks of nutrients, body
fluids, and anything that could be a nutrient source for maintenance of microorganism
biofilms in drains (23). Dispersion from a contaminated sink reservoir can result in
transmission to patients, either directly or indirectly mediated through numerous
contact surfaces. Herruzo and colleagues demonstrated the potential for microbial
transfer from contaminated hands, which continued to disperse microorganisms after
more than 10 successive contacts with surfaces (25).

The droplet dispersion load observed on settle plates was similar to and consistent
with our previous work (34). Total dispersion measured in corresponding experiments
in the previous study was higher, which may be attributed to one or more of the
following factors: (i) fewer settle plates were used in the present study (22 versus 90),
(ii) a higher water flow rate was used in the present study (8 versus 1.8 to 3.0 liters/min),
and (iii) air sampling methods performed in conjunction with the settle plate method
may have captured a portion of the dispersed droplets. Settle plates were found to be

January 2019 Volume 85 Issue 2 e01997-18

Characterization of Sink Microbial Particles Applied and Environmental Microbiology Downloaded from https://journals.asm.org/journal/aem on 26 March 2022 by 2a01:6500:a036:1f2b:3aae:6191:1bf:901.
aem.asm.org 7
a reliable method to assess the large-droplet dispersion from sinks. In this study, 22
settle plates (11.24 m2) were used, which accounted for a defined surface area and
locations on the sink counter. Dispersion could have been higher in locations on the
sink counter other than those chosen in the present study, and the dispersion load
recorded in this study may not be the absolute value. Of the three methods investi-
gated for air sampling, impaction and filtration were found to be reliable and consis-
tent. In the same amount of air sampled using impaction and filtration, comparable
counts were recorded; however, air sampled using the impinger method was unable to
capture the dispersion of GFP-E. coli under similar testing conditions.

Mannequin hands functioned as obstructions to the faucet water stream directly
impacting the sink drain, and therefore no dispersion was detected. This rationale
behind testing mannequin hands was to simulate hand washing, but in reality, the
water would be flowing before, after, and during a hand washing event. In other words,
an actual handwashing event is more dynamic than that of static mannequin hands and
is likely to have a direct impact on water on the sink drain at least for brief periods when
the water is running. There is also the scenario where the sinks and faucets may be used
outside of hand washing (e.g., dumping liquid wastes) (5, 8, 23). We think this finding
further defines and supports another important dynamic that may minimize dispersion
in health care settings (i.e., avoid faucet water flow directly onto drains to minimize
dispersion). All of these findings must be taken in the context of an experimental water
stream that directly hits the drain, which is outside of the Facility Guidelines Institute
(FGI) guidance but is thought to be frequently found in health care sink design. This
finding further supports the importance of enforcement of avoidance of this design in
a health care sink.

This study has several limitations. First, the dispersion experiments were not per-
formed in a controlled environment. Each dispersion experiment lasted at least 12 h,
and it was therefore not possible to maintain precisely the same conditions with regard
to airflow velocity, air temperature, relative humidity, and bacterial and/or fungal
burden in the laboratory space harboring the sinks. These parameters may have a direct
or indirect influence on the dispersion pattern and load recorded across experiments
(47). To address this issue, we monitored the heterotrophic plate counts, relative
humidity, and particle concentration in the air. Particle counts recorded in the absence
of a faucet event (control) were higher than or equal to that in the presence of a faucet
event (test). This observation implies that particle concentrations in the air were driven
by the relative humidity and/or temperature of the air. This trend was observed in all
the experimental methods (drain, P-trap inoculation, and drain colonization) (Fig. S1).
In other words, particle counts were largely consistent across the day for a given
experiment (control preceding test). Furthermore, the particle counter used in the
study could not resolve or measure particles of Ͼ5 ␮m, which defined droplet particles.
Another limitation was that air samples were collected at only one location relative to
the sink bowl, so it is not possible for this data set to define a “splash zone” pattern
without additional measurements collected from various positions and distance from
the source of dispersed organisms.

We have provided data to support the position that microorganisms will disperse
from contaminated sink bowl and drain surfaces primarily as large droplets that are
generated during faucet usage. These droplet-associated organisms remain viable, with
the potential to contaminate surfaces surrounding the sink bowl. However, with this
simulated sink design and study, it does not appear that dispersion results in the
production of bioaerosols with sustained dispersion characteristics that disperse into
the patient’s room space.

Based on the findings of this study and other contextual studies, we have some
further resolution around the dynamics of bacterial dispersion from sinks. A few
recommendations with regard to hospital sink design and usage practices should be
strictly implemented and practiced to minimize sink-related transmissions in hospitals,
namely, (i) faucet spouts should not flow directly into the drain; (ii) new sink designs
should limit counter space surrounding the sink bowl, and usage of counter space

January 2019 Volume 85 Issue 2 e01997-18

Kotay et al. Applied and Environmental Microbiology

FIG 4 Test sink, showing mannequin hands positioned directly below the faucet (A), and a peristaltic pump aem.asm.org 8 Downloaded from https://journals.asm.org/journal/aem on 26 March 2022 by 2a01:6500:a036:1f2b:3aae:6191:1bf:901.
on the right to deliver hand soap (B) via a steel tube attached to the faucet (C).

around existing sinks to place or store patient care items should be prohibited; (iii)
placement of sinks in patient care area should be strategically determined, keeping in
mind a 1-m droplet dispersion zone around the sinks; and (iv) patient sink usage should
be limited to hand washing, and disposal of nutrients or contaminated wastes into the
sinks should be prohibited.

MATERIALS AND METHODS

Sink gallery operation and automation. A dedicated sink gallery at the University of Virginia that
was described in an earlier study (34) was used in the present study. Sinks were operated in automated
mode using a microprocessor. The microprocessor activated the faucets for 30 s each hour via an inline
solenoid valve in the hot water supply line (faucet event) and also turned on a peristaltic pump
(Masterflex pump HV-77120-42; Cole-Parmer, Vernon Hills, IL) to dispense 1 ml of soap (Kleenex foam skin
cleaner; Kimberly-Clark Worldwide Inc., Roswell, GA). A steel tube discharging the soap was held in a
clamp attached to the faucet and positioned directly below the discharge of the faucet water. Water flow
rate from the faucet was 8 liters/min and the stream of water hit directly onto the drain. In one
experiment, mannequin hands (Dianne Practice Hand, D902; Fromm International, Mt. Prospect, IL)
attached to a metal rack and positioned between the faucet head and sink drain were used (Fig. 4),
preventing the stream of water from directly hitting the drain. To facilitate access to the luminal surface
of the drain line, sampling ports were drilled along the length of the tailpiece (between the P-trap and
the drain) and the trap arm (between the P-trap and the common line). These holes were fitted with size
00 silicone stoppers (Cole-Parmer, Vernon Hills, IL). Temperature and total and free chlorine residual
concentrations of the faucet water at (i) first catch and (ii) after 2 min flushing of faucets were measured
at regular intervals using the N,N-diethyl-p-phenylenediamine (DPD) method (Hach Model; Hach, Love-
land, CO).

Inoculation, growth, and establishment of GFP-E. coli colonies in sink P-traps. A single isolated
colony of GFP-E. coli (ATCC 25922GFP) cells grown from Ϫ80°C stock was inoculated in 5 ml tryptic soy
broth (TSB) containing 100 ␮g/ml ampicillin (ATCC medium 2855). The method of inoculation varied for
each experiment. For P-traps, a 10-ml mid-log-phase culture of GFP-E. coli (109 CFU/ml) was added into
the P-trap water (ϳ150 ml) through the lowermost sampling port on the tailpiece, using a 60-ml syringe
attached to silicone tubing (Cole-Parmer, Vernon Hills, IL). The inoculum was mixed with the P-trap water
by repeated withdrawal and injection of the inoculum, with precautions taken to avoid unintentional
inoculation of the drain (strainer) or sink bowl (bowl). For the drain inoculation, a 10-ml mid-log-phase
culture of GFP-E. coli (109 CFU/ml) was evenly applied on the surface of the drain using a sterile pipette.
For establishment of GFP-E. coli P-trap biofilm, a 10-ml mid-log-phase culture of GFP-E. coli (109 CFU/ml)
was added into an unused P-trap and, following inoculation, 25 ml TSB and 25 ml (ϫ2) 0.85% saline was
added on a daily basis through the drain for 7 days to facilitate biofilm growth on the luminal surface of
the drain line without additional water every hour. Seven days later, P-trap water and swab samples from
the inner surfaces of the drain, tailpiece, and the P-trap were plated on tryptic soy agar (TSA) containing
100 ␮g/ml ampicillin. TSA plates were incubated overnight at 37°C, and the CFU fluorescing under UV
light were enumerated. All preparatory culturing of GFP-E. coli cells took place in a separate room from
the sink gallery to avoid unintentional contamination.

January 2019 Volume 85 Issue 2 e01997-18

Characterization of Sink Microbial Particles Applied and Environmental Microbiology

FIG 5 Experimental set-up used for different air sampling methods. (a) Impaction, (b) impinger, and (c) gel filtration. aem.asm.org 9 Downloaded from https://journals.asm.org/journal/aem on 26 March 2022 by 2a01:6500:a036:1f2b:3aae:6191:1bf:901.
Air samples were collected at the initial faucet event (0 min) and every 30 min thereafter. Faucet events (faucet
activation) occurred at 0, 60, and 120 min under test conditions. Faucets were not activated in control experiments.

Sampling and enumeration of GFP-E. coli. To monitor the growth of GFP-E. coli within the
plumbing, sterile cotton swabs (Covidien, Mansfield, MA) presoaked in 0.85% sterile saline were inserted
through sampling ports, and biofilm samples were collected by turning the swab in a circular motion on
the inner surface (ϳ20 cm2). Sample swabs were pulse-vortexed in 3 ml saline, and serial dilutions were
plated on TSA. The entire surface area of the drain, the faucet aerator, and sink bowl surfaces were
sampled using environmental sponge wipes (sponge-stick with neutralizing buffer, 3M, St. Paul, MN),
using overlapping and multidirectional motions. The sponge wipes were expressed in 90 ml of
phosphate-buffered saline containing Tween 80 (0.02%) (PBST) using a stomacher (400 Circulator;
Seward Ltd., UK). The eluate was concentrated by centrifugation, and plate counts were performed on
TSA and R2A agar (Becton, Dickinson and Company, Franklin Lakes, NJ). TSA plates were incubated at
35°C for 48 h, and fluorescent CFU were enumerated; R2A plates were incubated at 25°C for 7 days, and
CFU were counted.

Experimental approach for dispersion studies. Each dispersion experiment comprised a 30-s
faucet event repeated three times at 60-min intervals (Fig. 5). Each experiment also comprised a control,
(without faucet events) followed by a test (with faucet events). During the control experiments,
automation was turned off to withhold the faucet events. Three sinks were tested concurrently but
staggered by a few minutes to account for the variability in dispersion driven by faucet water flow rate,
airflow dynamics in the room, contact angle, and wastewater drainage/water backup rate.

Sampling droplet dispersion. TSA settle plates were used to capture the droplet dispersion.
Numbered TSA plates were laid out radially around the sink bowl. A fixed layout and number of settle
plates around the sinks was used for each dispersal experiment (Fig. 6). The counter space of each sink
was thoroughly disinfected with Caviwipes-1 (Metrex Research, LLC, Orange, CA) prior to each experi-
ment. TSA plates were then positioned on the sink counter surrounding the sink bowl. Additional plates
were attached to the faucets, plexiglass partitions, and faucet handles using adhesive tape. Plates were
not placed in the sink bowl. TSA plates were also placed Ͼ3 m away from the sink as negative controls.
Lids of the TSA plates were removed only for the duration of the dispersal experiment. Dispersion per
defined area (CFU/cm2) for settle plates was determined by dividing the CFU counts from the TSA plate
by the surface area of the plate.

Air sampling and particle counts. Each dispersion experiment comprised the collection of air
samples at five separate time points, as follows: time t ϭ 0 (first faucet event), t ϭ 30 min after the first
faucet event, t ϭ 60 min (second faucet event), t ϭ 90 min, and t ϭ 120 min (third faucet event) (Fig. 5).
Individual sink sampling was staggered by a few minutes to provide time for air sampler installation and
sampling of each sink. A control experiment period, in which faucets were not activated for the entire
120 min, was also performed for each sink. Three air sampling methods were tested: impaction,
impingement, and filtration. For the impaction method, two SAS90 air samplers (Bioscience International,

January 2019 Volume 85 Issue 2 e01997-18

Kotay et al. Applied and Environmental Microbiology

FIG 6 Graphical representation depicting the layout of the settle plates positioned around the sink and used to aem.asm.org 10 Downloaded from https://journals.asm.org/journal/aem on 26 March 2022 by 2a01:6500:a036:1f2b:3aae:6191:1bf:901.
capture droplet dispersion.

Rockville, MD) containing one TSA plate and one R2A plate each were positioned 12 in. from the sink
bowl and set for a 300-liter sample (at 90 liters/min for 200 s) (Fig. 5a). TSA and R2A plates from each air
sampling event were incubated as described earlier. A gel filtration device (MD8 portable air sampler;
Sartorius AG, Goettingen, Germany) fitted with disposable gelatin filters (Sartorius AG, Goettingen,
Germany) was positioned 12 in. from the sink bowl and set for a 300-liter sample (at 100 liters/min for
180 s) (Fig. 5c). Gelatin filters were carefully overlaid on TSA plates, which were as already described.
Liquid impingers (Ace Glass Inc. Vineland, NJ) were autoclaved and filled with 20 ml sterile phosphate-
buffered saline (PBS) prior to each experiment. Each was connected via a flow meter (Cole-Parmer,
Vernon Hills, IL) and vacuum pump (Cole-Parmer, Vernon Hills, IL). The impinger was positioned 12 in.
from the sink bowl (set at 6 liters/min for 50 min) to collect a 300-liter air sample (Fig. 5b). In a biological
safety cabinet, the liquid from the impinger was transferred to a sterile tube, vortexed, and filtered
through 0.22-␮m membrane filters (Pall Laboratories, Port Washington, NY), and 5 ml duplicate samples
were plated on TSA and R2A plates. Fluorescent CFU were enumerated after TSA plates were incubated
at 35°C for 48 h and counted. R2A plates were incubated for 7 days at 25°C and counted. Plates from air
impaction samples and samples collected from liquid impingement were shipped via overnight courier
to the Centers for Disease Control and Prevention (CDC) laboratories for processing and counting. Gel
filtration plates were processed and counted at the University of Virginia. Paired with air sampling,
particles of 0.3, 0.5, 0.7, 1.0, 2.0, and 5.0 ␮m were measured using a particle counter (GT-526; Met One
Instruments, Inc., Grants Pass, OR) placed 12 in. from the sink bowl (Fig. 5). With a runtime of 660 s each,
3 successive runs of the particle counter were performed, with the first run coinciding with t ϭ 0. The
particle counter also recorded relative humidity and air temperature.

Verification of GFP-E. coli. Fluorescent colonies on TSA-ampicillin plates were counted under a
long-wavelength UV light source. To verify that fluorescent colonies were GFP-E. coli, two fluorescent
colonies from each sample were randomly selected and first screened on MacConkey II agar (BD, Franklin
Lakes, NJ). The MacConkey II plates were incubated at 35°C for 24 h, and lactose fermenters were isolated
on tryptic soy agar with 5% sheep blood (TSA II) (BD, Franklin Lakes, NJ) and incubated under the same
conditions. Once colonies were isolated, they were identified using matrix assisted laser desorption
ionization–time of flight mass spectrometry (MALDI-TOF MS; Bruker, Billerica, MA), or using the Vitek 2
system (bioMérieux, Durham, NC).

GFP-E. coli detection in sink plumbing and faucet water. Prior to each experiment, 500 ml of
first-catch faucet water and a 500-ml sample collected after 2 min of flushing were collected in sterile
bottles containing sodium thiosulfate (0.18 g/liter) for dechlorination. Aliquots of faucet water samples
were filtered through a 0.45-␮m membrane filter, placed on TSA, and incubated at 35°C for 48 h. This was
done to test for GFP-E. coli in the faucet water supplied to each sink and to ensure that cross
contamination of GFP-E. coli had not occurred. First-catch and 2-min flush samples were diluted, plated
on R2A, incubated at 25°C for 7 days, and counted. After the experiment was completed, samples were
collected and processed to detect GFP-E. coli in the sink plumbing. A 50-ml aliquot of the P-trap water
was collected using a 60-ml syringe connected to sterile silicone tubing, samples were thoroughly vortex
mixed, and 1-ml and 5-ml aliquots were filtered through 0.22-␮m membrane filters and placed on TSA
and R2A media, respectively. The sink P-trap and tailpiece were removed from the sink lab units, filled
with faucet water, plugged, and shipped by overnight courier to the CDC for analysis. Swab samples from
the tailpipe and P-trap and sponge wipes were processed as already described to recover and quantify
biofilm organisms. Samples were placed on TSA and R2A media and counted, as described above.

January 2019 Volume 85 Issue 2 e01997-18

Characterization of Sink Microbial Particles Applied and Environmental Microbiology

SUPPLEMENTAL MATERIAL
Supplemental material for this article may be found at https://doi.org/10.1128/AEM

.01997-18.
SUPPLEMENTAL FILE 1, PDF file, 0.3 MB.

ACKNOWLEDGMENTS
This research was performed under a Research Collaboration Agreement between

University of Virginia and the CDC. A.J.M. was supported in part by an IPA agreement
between the CDC and the University of Virginia School of Medicine (15IPA1508992).

The findings and conclusions in this presentation are those of the author(s) and do
not necessarily represent the views of the Centers for Disease Control and Prevention.

We thank Alexander Kallen for helpful discussions; James Matheson, Maria Burgos-
Garay, Amanda Lyons, and Terri Davy for assistance with sample processing; and Will
Guilford for design of the sink laboratory and automation.

REFERENCES beta-lactamase contamination for patients in the intensive care unit. J Downloaded from https://journals.asm.org/journal/aem on 26 March 2022 by 2a01:6500:a036:1f2b:3aae:6191:1bf:901.
Hosp Infect 87:126 –130. https://doi.org/10.1016/j.jhin.2014.02.013.
1. Brooke J. 2008. Pathogenic bacteria in sink exit drains. J Hosp Infect 13. Leitner E, Zarfel G, Luxner J, Herzog K, Pekard-Amenitsch S, Hoenigl M,
70:198 –199. https://doi.org/10.1016/j.jhin.2008.06.017. Valentin T, Feierl G, Grisold AJ, Hogenauer C, Sill H, Krause R, Zollner-
Schwetz I. 2015. Contaminated handwashing sinks as the source of a
2. Hota S, Hirji Z, Stockton K, Lemieux C, Dedier H, Wolfaardt G, Gardam M. clonal outbreak of KPC-2-producing Klebsiella oxytoca on a hematology
2009. Outbreak of multidrug-resistant Pseudomonas aeruginosa coloni- ward. Antimicrob Agents Chemother 59:714 –716. https://doi.org/10
zation and infection secondary to imperfect intensive care unit room .1128/AAC.04306-14.
design. Infect Control Hosp Epidemiol 30:25–33. https://doi.org/10.1086/ 14. Ambrogi V, Cavalie L, Mantion B, Ghiglia MJ, Cointault O, Dubois D, Prere
592700. MF, Levitzki N, Kamar N, Malavaud S. 2016. Transmission of metallo-
beta-lactamase-producing Pseudomonas aeruginosa in a nephrology-
3. La Forgia C, Franke J, Hacek D, Thomson R, Robicsek A, Peterson L. 2010. transplant intensive care unit with potential link to the environment. J
Management of a multidrug-resistant Acinetobacter baumannii outbreak Hosp Infect 92:27–29. https://doi.org/10.1016/j.jhin.2015.09.007.
in an intensive care unit using novel environmental disinfection: A 15. Chapuis A, Amoureux L, Bador J, Gavalas A, Siebor E, Chretien ML, Caillot
38-month report. American J Infection Control 38:259 –263. https://doi D, Janin M, de Curraize C, Neuwirth C. 2016. Outbreak of extended-
.org/10.1016/j.ajic.2009.07.012. spectrum beta-lactamase producing Enterobacter cloacae with high MICs
of quaternary ammonium compounds in a hematology ward associated
4. Breathnach AS, Cubbon MD, Karunaharan RN, Pope CF, Planche TD. with contaminated sinks. Front Microbiol 7:1070. https://doi.org/10
2012. Multidrug-resistant Pseudomonas aeruginosa outbreaks in two .3389/fmicb.2016.01070.
hospitals: association with contaminated hospital waste-water systems. 16. Clarivet B, Grau D, Jumas-Bilak E, Jean-Pierre H, Pantel A, Parer S, Lotthe
J Hosp Infect 82:19 –24. https://doi.org/10.1016/j.jhin.2012.06.007. A. 2016. Persisting transmission of carbapenemase-producing Klebsiella
pneumoniae due to an environmental reservoir in a university hospital,
5. Lowe C, Willey B, O’Shaughnessy A, Lee W, Lum M, Pike K, Larocque C, France, 2012 to 2014. Euro Surveill 21:piiϭ30213. https://doi.org/10
Dedier H, Dales L, Moore C, McGeer A. 2012. Outbreak of extended- .2807/1560-7917.ES.2016.21.17.30213.
spectrum beta-lactamase-producing Klebsiella oxytoca infections associ- 17. Stjärne Aspelund A, Sjöström K, Olsson Liljequist B, Mörgelin M, Me-
ated with contaminated handwashing sinks. Emerg Infect Dis 18: lander E, Påhlman LI. 2016. Acetic acid as a decontamination method for
1242–1247. https://doi.org/10.3201/eid1808.111268. sink drains in a nosocomial outbreak of metallo-beta-lactamase-
producing Pseudomonas aeruginosa. J Hosp Infect 94:13–20. https://doi
6. Starlander G, Melhus A. 2012. Minor outbreak of extended-spectrum .org/10.1016/j.jhin.2016.05.009.
beta-lactamase-producing Klebsiella pneumoniae in an intensive care 18. Swan JS, Deasy EC, Boyle MA, Russell RJ, O’Donnell MJ, Coleman DC.
unit due to a contaminated sink. J Hosp Infect 82:122–124. https://doi 2016. Elimination of biofilm and microbial contamination reservoirs in
.org/10.1016/j.jhin.2012.07.004. hospital washbasin U-bends by automated cleaning and disinfection
with electrochemically activated solutions. J Hosp Infect 94:169 –174.
7. Kotsanas D, Wijesooriya WR, Korman TM, Gillespie EE, Wright L, Snook K, https://doi.org/10.1016/j.jhin.2016.07.007.
Williams N, Bell JM, Li HY, Stuart RL. 2013. “Down the drain”: 19. Zhou Z, Hu B, Gao X, Bao R, Chen M, Li H. 2016. Sources of sporadic
carbapenem-resistant bacteria in intensive care unit patients and hand- Pseudomonas aeruginosa colonizations/infections in surgical ICUs: asso-
washing sinks. Med J Aust 198:267–269. https://doi.org/10.5694/mja12 ciation with contaminated sink trap. J Infect Chemother 22:450 – 455.
.11757. https://doi.org/10.1016/j.jiac.2016.03.016.
20. Amoureux L, Riedweg K, Chapuis A, Bador J, Siebor E, Pechinot A,
8. Roux D, Aubier B, Cochard H, Quentin R, van der Mee-Marquet N. 2013. Chretien M, de Curraize C, Neuwirth C. 2017. Nosocomial infections with
Contaminated sinks in intensive care units: an underestimated source of IMP-19-producing Pseudomonas aeruginosa linked to contaminated
extended-spectrum beta-lactamase-producing Enterobacteriaceae in the sinks, France. Emerg Infect Dis 23:304 –307. https://doi.org/10.3201/
patient environment. J Hosp Infect 85:106 –111. https://doi.org/10.1016/ eid2302.160649.
j.jhin.2013.07.006. 21. Baranovsky S, Jumas-Bilak E, Lotthe A, Marchandin H, Parer S, Hicheri Y,
Romano-Bertrand S. 2017. Tracking the spread routes of opportunistic
9. Tofteland S, Naseer U, Lislevand J, Sundsfjord A, Samuelsen O. 2013. A premise plumbing pathogens in a haematology unit with water points-
long-term low-frequency hospital outbreak of KPC-producing Klebsiella of-use protected by antimicrobial filters. J Hosp Infect 98:52–59. https://
pneumoniae involving intergenus plasmid diffusion and a persisting doi.org/10.1016/j.jhin.2017.07.028.
environmental reservoir. PLoS One 8:e59015. https://doi.org/10.1371/ 22. Bousquet A, van der Mee-Marquet N, Dubost C, Bigaillon C, Larréché S,
journal.pone.0059015. Bugier S, Surcouf C, Mérat S, Blanchard H, Mérens A. 2017. Outbreak of
CTX-M-15–producing Enterobacter cloacae associated with therapeutic
10. Vergara-López S, Domínguez MC, Conejo MC, Pascual Á, Rodríguez-Baño
J. 2013. Wastewater drainage system as an occult reservoir in a pro- aem.asm.org 11
tracted clonal outbreak due to metallo-beta-lactamase-producing Kleb-
siella oxytoca. Clin Microbiol Infect 19:E490 –E498. https://doi.org/10
.1111/1469-0691.12288.

11. Knoester M, de Boer M, Maarleveld J, Claas E, Bernards A, Jonge E, van
Dissel J, Veldkamp K. 2014. An integrated approach to control a pro-
longed outbreak of multidrug-resistant Pseudomonas aeruginosa in an
intensive care unit. Clin Microbiol Infect 20:O207–O215. https://doi.org/
10.1111/1469-0691.12372.

12. Wolf I, Bergervoet P, Sebens F, van den Oever H, Savelkoul P, van der
Zwet W. 2014. The sink as a correctable source of extended-spectrum

January 2019 Volume 85 Issue 2 e01997-18

Kotay et al. Applied and Environmental Microbiology

beds and syphons in an intensive care unit. Am J Infect Control 45: 34. Kotay S, Chai W, Guilford W, Barry K, Mathers AJ. 2017. Spread from the Downloaded from https://journals.asm.org/journal/aem on 26 March 2022 by 2a01:6500:a036:1f2b:3aae:6191:1bf:901.
1160 –1164. https://doi.org/10.1016/j.ajic.2017.04.010. sink to the patient: in situ study using green fluorescent protein (GFP)
23. De Geyter D, Blommaert L, Verbraeken N, Sevenois M, Huyghens L, expressing-Escherichia coli to model bacterial dispersion from hand
Martini H, Covens L, Pierard D, Wybo I. 2017. The sink as a potential washing sink trap reservoirs. Appl Environ Microbiol 83:e03327-16.
source of transmission of carbapenemase-producing Enterobacteriaceae
in the intensive care unit. Antimicrob Resist Infect Control 6:24. https:// 35. Schneider H, Geginat G, Hogardt M, Kramer A, Durken M, Schroten H,
doi.org/10.1186/s13756-017-0182-3. Tenenbaum T. 2012. Pseudomonas aeruginosa outbreak in a pediatric
24. Lalancette C, Charron D, Laferriere C, Dolce P, Deziel E, Prevost M, Bedard oncology care unit caused by an errant water jet into contaminated
E. 2017. Hospital drains as reservoirs of Pseudomonas aeruginosa: siphons. Pediatr Infect Dis J 31:648 – 650. https://doi.org/10.1097/INF
multiple-locus variable-number of tandem repeats analysis genotypes .0b013e31824d1a11.
recovered from faucets, sink surfaces and patients. Pathogens 6:E36.
https://doi.org/10.3390/pathogens6030036. 36. Fusch C, Pogorzelski D, Main C, Meyer C, el Helou S, Mertz D. 2015.
25. Herruzo R, Ruiz G, Vizcaino MJ, Rivas L, Perez-Blanco V, Sanchez M. 2017. Self-disinfecting sink drains reduce the Pseudomonas aeruginosa biobur-
Microbial competition in environmental nosocomial reservoirs and dif- den in a neonatal intensive care unit. Acta Paediatr 104:e344 – e349.
fusion capacity of OXA48-Klebsiella pneumoniae: potential impact on https://doi.org/10.1111/apa.13005.
patients and possible control methods. J Prev Med Hyg 58:E34 –E41.
26. Varin A, Valot B, Cholley P, Morel C, Thouverez M, Hocquet D, Bertrand 37. Pitten FA, Panzig B, Schröder G, Tietze K, Kramer A. 2001. Transmission
X. 2017. High prevalence and moderate diversity of Pseudomonas aerugi- of a multiresistant Pseudomonas aeruginosa strain at a German university
nosa in the U-bends of high-risk units in hospital. Int J Hyg Environ hospital. J Hospital Infection 47:125–130. https://doi.org/10.1053/jhin
Health 220:880 – 885. https://doi.org/10.1016/j.ijheh.2017.04.003. .2000.0880.
27. Deasy EC, Moloney EM, Boyle MA, Swan JS, Geoghegan DA, Brennan
GI, Fleming TE, O’Donnell MJ, Coleman DC. 2018. Minimizing micro- 38. Siegel JD, Rhinehart E, Jackson M, Chiarello L. 2007. 2007 guideline for
bial contamination risk simultaneously from multiple hospital wash- isolation precautions: preventing transmission of infectious agents in
basins by automated cleaning and disinfection of U-bends with health care settings. Am J Infect Control 35:S65–164. https://doi.org/10
electrochemically activated solutions. J Hosp Infect 100:e98 – e104. .1016/j.ajic.2007.10.007.
https://doi.org/10.1016/j.jhin.2018.01.012.
28. Dewi R, Tarini A, Sukrama D. 2013. Phenotypic detection of 39. WHO. 2014. Infection prevention and control of epidemic- and
carbapenemase-producing gram-negative bacteria in hospital environ- pandemic-prone acute respiratory infections in health care: WHO guide-
ment. Abstr ASID Gram Negative “Superbugs” Meeting, Queensland, lines. World Health Organization, Geneva, Switzerland. http://www.who
Australia. .int/csr/bioriskreduction/infection_control/publication/en/.
29. Loveday H, Wilson J, Kerr K, Pitchers R, Walker J, Browne J. 2014.
Association between healthcare water systems and Pseudomonas 40. Kramer A, Daeschlein G, Niesytto C, Sissoko B, Sütterlin R, Blaschke M,
aeruginosa infections: a rapid systematic review. J Hosp Infect 86:7–15. Fusch C. 2005. Contamination of sinks and emission of nosocomial
https://doi.org/10.1016/j.jhin.2013.09.010. gramnegative pathogens in a NICU— outing of a reservoir as risk factor
30. Bloomfield S, Exner M, Flemming HC, Goroncy-Bermes P, Hartemann P, for nosocomial colonization and infection. Umweltmed Forsch Prax 10.
Heeg P, Ilschner C, Krämer I, Merkens W, Oltmanns P, Rotter M, Rutala
WA, Sonntag HG, Trautmann M. 2015. Lesser-known or hidden reservoirs 41. Doring G, Horz M, Ortelt J, Grupp H, Wolz C. 1993. Molecular epidemi-
of infection and implications for adequate prevention strategies: Where ology of Pseudomonas aeruginosa in an intensive care unit. Epidemiol
to look and what to look for. GMS Hyg Infect Control 10:Doc04. https:// Infect 110:427– 436. https://doi.org/10.1017/S0950268800050858.
doi.org/10.3205/dgkh000247.
31. Guyot A, Turton JF, Garner D. 2013. Outbreak of Stenotrophomonas 42. Zemouri C, de Soet H, Crielaard W, Laheij A. 2017. A scoping review on
maltophilia on an intensive care unit. J Hosp Infect 85:303–307. https:// bio-aerosols in healthcare and the dental environment. PLoS One 12:
doi.org/10.1016/j.jhin.2013.09.007. e0178007. https://doi.org/10.1371/journal.pone.0178007.
32. Parkes LO, Hota SS. 2018. Sink-related outbreaks and mitigation strate-
gies in healthcare facilities. Curr Infect Dis Rep 20:42. https://doi.org/10 43. Haig C, Mackay W, Walker J, Williams C. 2016. Bioaerosol sampling:
.1007/s11908-018-0648-3. sampling mechanisms, bioefficiency and field studies. J Hosp Infect
33. Gordon A, Mathers A, Cheong E, Gottlieb T, Kotay S, Walker A, Peto T, 93:242–255. https://doi.org/10.1016/j.jhin.2016.03.017.
Crook D, Stoesser N. 2017. The hospital water environment as a reservoir
for carbapenem-resistant organisms causing hospital-acquired infec- 44. Morawska L. 2006. Droplet fate in indoor environments, or can we
tions—a systematic review of the literature. Clin Infect Dis 64: prevent the spread of infection? Indoor Air 16:335–347. https://doi.org/
1435–1444. https://doi.org/10.1093/cid/cix132. 10.1111/j.1600-0668.2006.00432.x.

45. Eames I, Shoaib D, Klettner C, Taban V. 2009. Movement of airborne
contaminants in a hospital isolation room. J R Soc Interface 6:S757–S766.
https://doi.org/10.1098/rsif.2009.0319.focus.

46. Stetzenbach L. 2002. Introduction to aerobiology. In Hurst C, Crawford R,
Knudsen G, McInerney M, Stetzenbach L (ed), Manual of environmental
microbiology, 2nd ed. ASM Press, Washington, DC.

47. Mohr A. 2002. Fate and transport of microorganisms in the air, p
827– 838. In Hurst C, Crawford R, Knudsen G, McInerney M, Stetzenbach
L (ed), Manual of environmental microbiology, 2nd ed. ASM Press,
Washington, DC.

January 2019 Volume 85 Issue 2 e01997-18 aem.asm.org 12

Annals of Burns and Fire Disasters - vol. XXXII - n. 1 - March 2019

EXTENSIVELY DRUG-RESISTANT PSEUDOMONAS AERUG-
INOSA OUTBREAK IN A BURN UNIT: MANAGEMENT AND
SOLUTIONS

ÉPIDÉMIE À PSEUDOMONAS ÆRUGINOSA DANS UN CTB: ÉVALUATION
ET ÉRADICATION

Aguilera-Sáez J.,* Andreu-Solà V., Larrosa Escartín N., Rodríguez Garrido V., Armadans Gil L.,
Sánchez García J.M., Campins M., Baena Caparrós J., Barret J.P.

Vall d’Hebron University Hospital, Barcelona, Spain

SUMMARY. Infections are still the main cause of mortality in burn patients. Multidrug resistant bacteria
can cause outbreaks in critical care and burn units. We describe an outbreak of infection by extensively
drug-resistant Pseudomonas aeruginosa in the Burn Unit of a University Hospital in Barcelona (Spain) be-
tween April and July 2016. A descriptive study of all cases, a bacterial colonization screening of all admitted
patients and a microbiological environmental study were performed in order to detect a possible common
focus. Contact isolation and cohortization of healthcare workers of all infected or colonized patients were
applied. Environmental control measures were instituted for possible sources of infection. The outbreak was
caused by a strain of P. aeruginosa only sensitive to colistin. Ten patients were infected or colonized and
two of them died. The same strain was detected in several taps and drains in different rooms of the Unit.
After applying control measures, changing faucets and drains, carrying out thermal disinfection of the hot
water installation of the unit, disinfecting the rooms with ultraviolet radiation and placing antibacterial fil-
tration devices in all the taps among other measures, an effective control of the outbreak was achieved.

Keywords: burns, Pseudomonas aeruginosa, outbreak, burn centre, extensively drug-resistant

RÉSUMÉ. Les infections sont toujours une cause majeure de mortalité chez les brûlés. Des épidémies à
bactéries multirésistantes (BMR) dans les CTB sont régulièrement rapportées. Nous décrivons une épidémie
due à Pseudomonas æruginosa BMR, sensible uniquement à la colimycine, survenue dans le CTB d’un hô-
pital universitaire de Barcelone entre avril et juillet 2016. Elle a touché 10 patients dont 2 sont morts. Une
étude de chaque cas, un dépistage chez tous les entrants et une étude environnementale ont été réalisées,
afin de trouver d’éventuelles similitudes. Un isolement contact et un cohorting ont été mis en place. Des
mesures de contrôle de l’environnement ont été implémentées. La souche incriminée a été retrouvée dans
plusieurs robinets et siphons du service. Cette épidémie a été résolue après, outre les mesures précitées,
changement des robinets et des siphons (avec mise en place d’ultrafiltres sur les robinets), choc thermique
du réseau d’adduction d’eau, désinfection terminale UV des chambres.

Mots-clés : brûlure, Pseudomonas æruginosa, épidémie, CTB, BMR

___________

* Corresponding author: J. Aguilera-Sáez, Department of Plastic Surgery and Burn Centre, Vall d’Hebron University Hospital, c/ Passeig de la Vall d’Hebron 119-129,
08035, Barcelona, Spain. Tel.: +34 934893475; email: [email protected]
Manuscript: submitted 14/03/2019, accepted 21/04/2019

47

Annals of Burns and Fire Disasters - vol. XXXII - n. 1 - March 2019

Introduction tients. It also has its own operating room, a room for
outpatient treatment, an emergency room, a rehabili-
After the initial resuscitation of burned patients, up tation area, a room for hydrotherapy and a games
to 75% of patient deaths are related to an infectious room, along with other rooms for storage and work
process.1 Several factors contribute to the high fre- areas for healthcare workers. It forms part of the Cen-
quency and severity of these infections and to their tres, Services and Reference Units (CSUR) of the
multiple possible locations: the destruction of the skin Spanish National Health System for critically burned
allows easy access of microorganisms, the presence of patients, and its direct area of influence includes Cat-
necrotic tissue provides an optimal growth medium alonia, the Balearic Islands and Andorra, covering a
for them, invasive monitoring facilitates other entry population of more than 8,000,000 people.
points and, finally, nonspecific alteration of immune
function allows microbial proliferation.1 Epidemiological investigation
A case was defined as any patient admitted to the
Pseudomonas aeruginosa is a gram-negative, Burn Unit who, as of April 2016, presented infection
aerobic and non-fermenting bacterium that is widely or colonization by P. aeruginosa with “in vitro” re-
distributed in nature. Soil, plants or running water sistance to all the antibiotics evaluated except col-
can act as a reservoir, with a predilection for humid, istin. Weekly screening by rectal smear was
even nutrient-deficient environments.2,3 The ease performed on all patients admitted to the Unit to de-
with which P. aeruginosa grows both in nature and tect possible colonization, in addition to the cultures
inside hospitals, and its capacity to acquire resist- that were performed in case of suspected infection.
ance mechanisms to antibiotics, make it one of the This screening was extended to patients who went
most significant causes of serious nosocomial infec- to the outpatient and rehabilitation treatment rooms
tion, affecting mainly immunocompromised pa- and who had been admitted as of April 2016.
tients.4,5 Infection between hospitalized patients
occurs mainly through the hands of healthcare work- Microbiological study
ers but also through contact with contaminated sur- In order to detect possible reservoirs and environ-
faces and objects.3 mental sources from which the pathogen could have
been transmitted, seven rounds of environmental sam-
The number of published outbreaks caused by pling were performed in which samples were taken
multidrug-resistant (MDR) bacteria in burn units with a swab, and on three occasions, water samples
over the last 30 years is relatively small. Of these, were taken. A total of 290 samples were collected,
few come from developed countries, and diagnostic which included hot and cold water, moisturizers, soaps
and screening techniques have changed in recent and antiseptic solutions, the aerators from all taps, the
years.6 Early detection of the outbreak can facilitate internal surfaces of the tap pipes and all of the drain
its control and limit its spread. The objective of this grates, the surfaces of containers and other environ-
work is to describe an outbreak caused by exten- mental surfaces (Table I). The surfaces were sampled
sively drug-resistant P. aeruginosa (XDR-PA)7 in the using wet sponges with neutralizing buffer (3MTM
Burn Unit of the Vall d’Hebron University Hospital Sponge-Stick, USA). Likewise, samples of hot and
in Barcelona (Spain) between April and July 2016, cold water were obtained from all of the taps in the
and the measures adopted to control it. Unit. The samples of hot water were obtained follow-
ing the recommendations published in “Health Tech-
Material and methods nical Memorandum 04-01: Safe water in healthcare
premises: Part C - Pseudomonas aeruginosa - advice
The Vall d’Hebron University Hospital of for augmented care units”. In the hot water taps, a 2-L
Barcelona (Spain) is a health centre with more than water sample was obtained after 2 hours of not using
1100 beds. The Burn Unit is an integral unit that has the tap, and another 2-L sample was obtained after 2
six beds for critical patients, 16 beds for semi-critical min of letting the water circulate.
or non-critical patients and four beds for pediatric pa-

48

Annals of Burns and Fire Disasters - vol. XXXII - n. 1 - March 2019

The sponges were incubated overnight at 37ºC and and the analysis of the macro-restriction profiles was
were then planted in selective and differential media to performed with the GelCompar II program (Applied
investigate the presence of P. aeruginosa and mul- Maths, Belgium) using the Dice similarity index,
tidrug-resistant gram-negative bacilli. The water sam- which is based on UPGMA (unweighted pair group
ples were filtered through a 0.45-µm Millipore filter method using arithmetic averages). For the genera-
and cultured in the appropriate selective media. tion of the dendrograms, a position tolerance value
of 1.0% was accepted. The clonal relationship was
The clinical samples were seeded in selective and determined by comparisons of the macrorestriction
differential culture media, some of which specific for profiles of each of the isolates studied. Thus, two
the search for resistant microorganisms. In the case of isolates belonging to the same clone were defined
urine and respiratory samples, quantitative cultures based on a similarity higher than 85%.10
were inoculated. All media were incubated at 35ºC in
suitable atmospheres, with reading and interpretation Control measures taken
after at least 24 hours had elapsed. 1 All infected or colonized patients were sub-

As regards monitoring samples (stool samples taken jected to contact precautions. Isolation pre-
by rectal swab) for colonization, active searches for cautions were adopted by the healthcare
multidrug-resistant gram-negative bacilli were made workers (personnel dedicated exclusively to
with specific selective media. infected or non-infected patients).
2 Compliance with basic infection prevention
Antimicrobial susceptibility to beta-lactam antibi- measures (hand hygiene, aseptic measures,
otics with antipseudomonal activity (piperacillin- cleaning and disinfection of surfaces and
tazobactam, aztreonam, ceftazidime, cefepime, materials) was reinforced through training
imipenem and meropenem), aminoglycosides (gentam- sessions with the professionals of the unit.
icin, tobramicin and amikacin), quinolones 3 A review of the antimicrobial treatment pro-
(ciprofloxacin), fosfomycin and colistin was deter- tocols and antibiotic policy of the unit was
mined by disk diffusion in Mueller-Hinton agar and an- conducted, with the aim of reducing antibi-
tibiotic gradient strip test, following the EUCAST otic pressure and limiting the development
guidelines (www.eucast.org). The presence of metallo- of multidrug-resistance. We do not perform
beta-lactamase production (MBL) was screened by the empirical systemic antibiotic treatment in a
modified Hodges test and double-disk synergy test generalized way on admission. In cases
using imipenem and imipenem+EDTA and confirmed where infection is suspected in the first 72
by Polymerase Chain Reaction (PCR). In this case spe- hours of admission, in patients who have had
cific primers for the detection of the MBL-encoding no previous relationship with the health sys-
genes blaVIM, blaIMP and blaNDM were used.8 tem and no colonization by Pseudomonas
aeruginosa is suspected, amoxycillin/clavu-
Clonal relatedness lanic acid is preferentially administered, after
obtaining samples for microbiological cul-
Pulsed-field gel electrophoresis (PFGE) of the re- ture. In contrast, in patients with criteria of
striction fragments of total bacterial DNA, obtained possible colonization or infection by
by means of the restriction enzyme Spel (Roche, Pseudomonas aeruginosa or with more than
USA) was performed to confirm the clonal relation- 72 hours of admission, broad-spectrum an-
ships of the P. aeruginosa isolates. All of the isolates tibiotic therapy is initiated, preferably
from the culture of the clinical samples, along with piperacillin-tazobactam. The use of car-
those from the previously mentioned colonization bapenems, aminoglycosides, glycopeptides
and environmental samples, were studied using this or other antimicrobials is reserved for pa-
technique. DNA separation was performed in the tients with septic shock or prolonged admis-
CHEF-DRII pulsed field electrophoresis system
(Bio-Rad, USA) according to the conditions de-
scribed previously.9 The gel images were captured
using the Gel DocTM XR system (Bio-Rad, USA),

49

Annals of Burns and Fire Disasters - vol. XXXII - n. 1 - March 2019 Results
8
4
P. aeruginosa
II
8
Table

Table II Fig. 1
P. aeruginosa

Fig. 1

Fig. 1

Annals of Burns and Fire Disasters - vol. XXXII - n. 1 - March 2019

progressed to multi-organ failure and death; in one negative for XDR-PA (Fig. 1); therefore, the out-
of them, the septic shock was secondary to a respi- break was considered controlled in September 2016.
ratory infection by XDR-PA, and in the other, it was Patient no. 8 remained hospitalized and colonized
secondary to an infection of the burns by XDR-PA. by XDR-PA until October 2016. For the early detec-
tion of a possible recontamination, a protocol was
The results of the environmental study are shown established that includes the monthly sampling of
in Table I. The strain of XDR-PA that caused the the Unit’s water and the taps and plumbing. On 1st
outbreak was detected in the tap aerator and the tap November 2018, no new episodes of XDR-PA infec-
plumbing of four rooms occupied with patients col- tion had been detected, nor were positive environ-
onized or infected by said bacteria. This strain was mental samples detected.
also detected in the grates of the basin drains of two
rooms occupied by infected patients (in one of these Discussion
rooms, two consecutive patients acquired XDR-PA;
in the drain of the basin in the other room, XDR-PA P. aeruginosa is an opportunistic bacterium with
was detected only in the second sampling), in the a large environmental reservoir that can cause seri-
aerator of the tap of the hydrotherapy room and in ous systemic infections of any type (respiratory, uri-
the grate of the basin drain of the nurses’ station nary, cutaneous or soft-tissue infections, or
(Fig. 2). bacteremia, among others) in patients with severe
comorbidities or some type of immunosuppression,
The molecular epidemiology study using the such as occurs with extensive burns. Its multiple
PFGE technique showed that all the isolates of structural (capsular exopolysaccharides, adhesins,
XDR-PA obtained from clinical and environmental pili, diffusible pigments, endotoxins), toxigenic (ex-
samples were identical. otoxins A, S and T) and enzymatic (elastase, alkaline

With the progressive application of the control
measures mentioned above, no more cases were de-
tected, and all of the environmental samples were

Table I - Summary of environmental studies, and objects from which the outbreak strain was isolated

Environmental Date Objects investigated (n samples) Pseudomonas aeruginosa positive MDR Pseudomonas
study samples (n samples) aeruginosa positive objects

1 14/06/2016 Faucet aerators (7) 4 faucet aerators 4 faucet aerators
Sink drain grilles (7) 1 sink drain grille 1 sink drain grille

2 21/06/2016 Tap pipes (6) 4 tap pipes 4 tap pipes

3 27/06/2016 Hot water (12 samples) 0 0

4 12/07/2016 Faucet aerators (5) 1 sink drain grille 1 sink drain grille
Sink drain grilles (3)

5 22/07/2016 Hot water (92 samples) Hot water (1 sample) 1 faucet aerator

Faucet aerators (48) Faucet aerators (5) 1 sink drain grille
Sink drain grilles (45) Sink drain grilles (19)

Environmental surfaces (18) Environmental surfaces (3)

Body milks, soaps & antiseptics (9) Body milks, soaps & antiseptics (1)

Container surfaces (6)

6 29/07/2016 Cold water (5) 0 0

7 23/08/2016 Faucet aerators (3) 4 sink drain grilles 1 sink drain grille
Sink drain grilles (26)
Environmental surfaces (1)

51

Annals of Burns and Fire Disasters - vol. XXXII - n. 1 - March 2019

P. aeruginosa

Fig. 2

Table II

Case Age Date of First positive Sample TBSA (%) Mechanical Urinary Total Evolution
admission culture date ventilation catheter admission
Discharged
1 80-90 23/03/2016 25/04/2016 Urine 15% No Yes days Discharged
2 30-40 28/04/2016 05/05/2016 Urine 30% Yes Yes 43 Discharged
3 50-60 16/04/2016 09/05/2016 Rectal swab 21% No Yes 63 Discharged
4 40-50 02/05/2016 09/05/2016 Rectal swab 28% No Yes 38 Discharged
5 70-80 11/04/2016 17/05/2016 Urine Sequels of No No 7
scleroderma 50 Discharged
6 40-50 10/05/2016 27/05/2016 Rectal swab 5.5 % No No Dead
7 50-60 24/04/2016 28/05/2016 Tracheal aspirate 40% Yes Yes 23
8 20-30 12/05/2016 02/06/2016 70% Yes Yes 57 Admitted
9 30-40 29/06/2016 06/07/2016 Wound swab 80% Yes Yes .. Dead
10 40-50 08/07/2016 18/07/2016 Tracheal aspirate 88% Yes Yes 39
Tracheal aspirate 62 Discharged

Annals of Burns and Fire Disasters - vol. XXXII - n. 1 - March 2019

strated to be one of the most frequent causes of ac- quate to relate the isolates,3 such as PFGE, as used
quiring P. aeruginosa in Intensive Care Units (ICU) in the present outbreak.
since only between 20% and 33% of patients are col-
onized at admission.3 In addition, cross-transmission The transmission of strains from patients to drains
between patients has been identified in between 8% has also been described in the literature.13 This ob-
and 50% of infections or colonizations acquired dur- servation emphasizes the importance of minimizing
ing admission.3 In these cases, healthcare workers room changes of patients infected or colonized by
would be the most likely vector. Reinforcing hand XDR-PA as much as possible to avoid the spread of
hygiene of healthcare workers is one of the most es- these microorganisms. For all of these reasons, the
sential measures to prevent cross-transmission of in- disinfection of the taps, basins and drains of the ICUs
fections. An educational program to promote hand and Burn Units is of vital importance. However, Gar-
hygiene among healthcare workers is well estab- vey et al.15 demonstrated that following the national
lished in our centre following the World Health Or- guidelines of the British Department of Health on the
ganization (WHO) multimodal hand hygiene use and maintenance of the water supply system was
strategy12 and annual monitoring of hand hygiene not sufficient since transmission persisted. This as-
compliance is performed by direct observational sessment has led to the description of alternative
method. Alcohol-based handrub is the most used methods for cleaning and disinfecting the surfaces of
method for routine hand antisepsis in our setting, the basins, grates and drains.18,19 Finally, it has been
using soap and water when hands are visibly dirty or described that the use of point-of-use tap filters is as-
soiled with blood or body fluids. In 2016, the ob- sociated with a decrease in number of colonizations
served adherence of hand hygiene in the burn unit of and infections,3,20 with a lower cost than a strategy
our hospital was about 70.4%. involving the use of sterile bottled water.20

However, for 30% to 60% of the remaining cases, However, in our outbreak, after having identified
there is plausible evidence from previous studies that the facility that supplies sanitary hot water as a po-
strains of P. aeruginosa found in washbasins, hy- tential source and having adopted corrective meas-
drotherapy tanks, taps and water from these taps can ures (replacement of taps and aerators, disinfection),
be a source for colonization and infection of patients, two more cases of XDR-PA infection were detected.
especially in ICUs.3,13-15 The persistence of P. aerug- In addition, the times between admission and acqui-
inosa in adverse media is due to its capacity to form sition of XDR-PA were shorter in these two cases
biofilms.5 Transmission might be caused by splash- compared with the others. These two cases corre-
ing of water droplets from contaminated sinks to sponded to patients who were admitted to rooms in
healthcare personnel’s hands, which may become which other infected patients had been admitted. In
transiently colonized;16 then these hands might come a previous prospective observational study,21 it was
into contact with the patient or with invasive devices found that admission to an ICU room previously oc-
such as an orotracheal tube, a vascular catheter or a cupied by a patient with MDR-PA increases the risk
urinary catheter. Although in outbreaks related to of acquisition of this bacterium in subsequent pa-
water environment the routes of transmission have tients, and the time to acquire the bacterium was also
been difficult to characterize, in most of them the re- shorter. This has been related to the fact that disin-
inforcement of hand hygiene and contact precautions fection and cleaning of these rooms does not com-
has been associated with a significant reduction in pletely eliminate these pathogens22 and suggests that
transmission.17 In the present outbreak, XDR-PA was the contamination of surfaces of the rooms, and of
found in the tap aerators of four rooms and in the the basins and drains, can play a role in the transmis-
grates of the basin drains of two additional rooms sion of MDR-PA.21 Interestingly, in the present study,
where patients colonized or infected by XDR-PA XDR-PA was detected in the drain grate of a basin
were admitted. To relate the possible sources with the from one of the two rooms where two patients were
cases, an epidemiological study is required, and cur- positive for XDR-PA. Another recent study23 associ-
rently only genotypic methods are considered ade- ated a more intense terminal disinfection of the
rooms with a decrease in the risk of infection of the

53

Annals of Burns and Fire Disasters - vol. XXXII - n. 1 - March 2019

organisms studied (methicillin-resistant Staphylo- measures) could be justified if the eradication of the
coccus aureus, vancomycin-resistant Enterococcus, outbreak and a decrease in the number of baseline
and multidrug-resistant Clostridium difficile and cases with multidrug-resistant microorganisms are
Acinetobacter spp.); however, this work did not in- achieved.
clude MDR-PA as an object of study. This observa-
tion suggests that objects and surfaces may play a On February 27, 2017, WHO published a list of
role in the spread of certain microorganisms and, the 12 bacteria that require priority research for the
therefore, that the use of the ultraviolet radiation dis- development of new antibiotics.28 P. aeruginosa re-
infection system used in the present outbreak may sistant to carbapenems is one of the critical priority
have played an important role in their control. group. These multidrug-resistant bacteria constitute
a serious public health problem.
The investigation of the survival mechanisms of
inanimate environments and the dispersion mecha- Conclusions
nisms of these microorganisms is fundamental to un-
derstanding and stopping these outbreaks. For In April 2016, there was an outbreak of infection
example, it would be useful to investigate the de- by extensively drug-resistant P. aeruginosa in the
signs of taps and drains to reduce their contamina- Burn Unit of a university hospital in Barcelona
tion and to avoid spreading, since one of the (Spain), with 10 patients affected. The epidemiolog-
postulated forms of transmission is related to splash- ical investigation revealed the existence of environ-
ing from the drains, possibly due to basins designed mental reservoirs in the taps and drains of the Unit.
in such a way that the stream of water directly hits The application of strict control measures against
the drain.19 transmission, the strict adherence by healthcare
workers to infection control policy, the use of an an-
Several studies have been published on the im- tibiotic protocol, along with the replacement of taps
pact that multidrug-resistant P. aeruginosa has on and drains and the installation of bacterial filters
the increase in mortality rates, the hospitalization pe- were associated with an effective control of the out-
riod and, to a lesser extent, the economics. In gen- break.
eral, it is estimated that infections caused by
multidrug-resistant P. aeruginosa are associated with P. aeruginosa infections continue to be a major
a higher mortality rate, a prolonged period of hospi- concern among burn patients because they drasti-
talization and increased costs compared with infec- cally worsen the prognosis of these patients, and out-
tions caused by antibiotic-susceptible bacteria, breaks are not uncommon in burn units. We believe
especially in terms of increased pharmaceutical that research should be promoted in this field, as
costs.24-27 The secondary expense of the implemen- should the exchange of experiences to control one
tation of the different control measures (replacement of the greatest threats to modern medicine: microor-
of contaminated taps and drains, bacterial filtration ganisms resistant to multiple antibiotics.
systems in taps, ultraviolet irradiation, among other

BIBLIOGRAPHY 4 Kasper DL, Braunwald E, Fauci AS, Hauser SL et al.: Harrison’s
Principles of Internal Medicine, 17th ed., 202-208, New York, 2008.
1 Arsemino M, Hemsley C: ABC of burns: Intensive care management
and control of infection. BMJ Br Med J, 239(7459): 220-3, 2004. 5 Moradali MF, Ghods S, Rehm BH: Pseudomonas aeruginosa
lifestyle: a paradigm for adaptation, survival, and persistence. Cell
2 Montero MM: Pseudomonas aeruginosa multiresistente: aspectos Frontal Infect Microbiol, 7: 39, 2017.
epidemiológicos, clínicos y terapêuticos [dissertation], 109. Bella-
terra, Universitat Autònoma de Barcelona, 2013. Available from: 6 Girerd-Genessay I, Benet T, Vanhems P: Multidrug-resistant bacterial
http//hdl.handle.net/10803/107902. outbreaks in Burn Units: a synthesis of the literature according to the
ORION statement. J Burn Care Res, 37(3): 172-80, 2016.
3 Trautmann M, Lepper PM, Haller M: Ecology of Pseudomonas
aeruginosa in the intensive care unit and the evolving role of water 7 MagiorakosAP, SrinivasanA, Carey RB, Carmeli Y et al.: Multidrug-
outlets as a reservoir of the organism. Am J Infect Control, 33(5 resistant, extensively drug-resistant and pandrug-resistant bacteria:
SUPPL. 1): 41-9, 2005. an international expert proposal for interim standard definitions for
acquired resistance: international standard definitions for acquired
54

Annals of Burns and Fire Disasters - vol. XXXII - n. 1 - March 2019

resistance. Clin Microbiol Infecti, 18(3): 268-81, 2012. hand basin tap disinfection. J Hosp Infect, 94(1): 21-2, 2016.
8 Tortola MT, Lavilla S, Miró E, González JJ et al.: First detection of 19 Stjärne Aspelund A, Sjöström K, Olsson Liljequist B, Mörgelin M

a carbapenem-hydrolyzing metalloenzyme in two Enterobacteriaceae et al.: Acetic acid as a decontamination method for sink drains in a
isolates in Spain. Antimicrob Agents Chemother, 49: 3492-4, 2005. nosocomial outbreak of metallo-β-lactamase-producing
9 Kaufmann ME: Pulsed-field gel electrophoresis. Methods Mol Med, Pseudomonas aeruginosa. J Hosp Infect, 94(1): 13-20, 2016.
15: 33-50, 1998. 20 Hall J, Hodgson G, Kerr KG: Provision of safe potable water for im-
10 Tenover FC, Arbeit RD, Goering RV, Mickelsen PA et al.: Interpret- mune-compromised patients in hospital. J Hosp Infect, 58(2): 155-
ing chromosomal DNA restriction patterns produced by pulsed-field 8, 2004.
gel electrophoresis: criteria for bacterial strain typing. J Clin Micro- 21 Nseir S, Blazejewski C, Lubret R, Wallet F et al.: Risk of acquiring
biol, 33: 2233-9, 1995. multidrug-resistant gram-negative bacilli from prior room occupants
11 Montero M, Horcajada JP, Sorlí L, Alvarez-Lezma F et al.: Effec- in the intensive care unit. Clin Microbiol Infect, 17: 1201-1208, 2011.
tiveness and safety of colistin for the treatment of multidrug-resistant 22 Otter JA: Interpreting the Joint Working Party Guidelines for cleaning
Pseudomonas aeruginosa infections. Infection, 37(5): 461-5, 2009. up after carbapenemase-producing organisms: the devil’s in the di-
12 World Health Organization: Patient safety a world alliance for safer lution. J Hosp Infect, 94: 108-109, 2016.
health care. WHO guidelines on hand hygiene in health care. First 23 Anderson DJ, Chen LF, Weber DJ, Moehring RW et al.: Enhanced
global patient safety challenge clean care is safer care. WHO, terminal room disinfection and acquisition and infection caused by
Geneva, 2009. multidrug-resistant organisms and Clostridium difficile: a cluster-
13 Tredget EE, Shankowsky HA, Joffe AM, Inkson TI et al.: Epidemi- randomized, multicentre, crossover study. Lancet, 389: 805-814,
ology of infections with Pseudomonas aeruginosa in burn patients: 2017.
the role of hydrotherapy. Clin Infect Dis, 15: 941-9, 1992. 24 Bou R, Lorente L, Aguilar A, Perpiñán J et al.: Hospital economic
14 Reuter S, Sigge A, Wiedeck H, Trautmann M: Analysis of transmis- impact of an outbreak of Pseudomonas aeruginosa infections. J Hosp
sion pathways of Pseudomonas aeruginosa between patients and tap Infect, 71: 138-142, 2009.
water outlets. Crit Care Med, 30: 2222-8, 2002. 25 Cosgrove SE: The relationship between antimicrobial resistance and
15 Garvey MI, Bradley CW, Tracey J, Oppenheim B: Continued trans- patient outcomes: mortality, length of hospital stay, and health care
mission of Pseudomonas aeruginosa from a wash hand basin tap in costs. Clin Infect Dis, 42: S82-S89, 2006.
a critical care unit. J Hosp Infect, 94(1): 8-12, 2016. 26 Morales E, Cots F, Sala M, Comas M et al.: Hospital costs of noso-
16 Kanamori H, Weber DJ, Rutala WA: Healthcare outbreaks associated comial multi-drug resistant Pseudomonas aeruginosa acquisition.
with a water reservoir and infection prevention strategies. Clin Infect BMC Health Services Research, 12: 122, 2012.
Dis, 62(11): 1423-1435, 2016. 27 Carmeli Y, Troillet N, Karchmer AW, Samore MH: Health and eco-
17 Kizny Gordon AE, Mathers AJ, Cheong EYL, Gottlieb T et al.: The nomic outcomes of antibiotic resistance in Pseudomonas aeruginosa.
hospital water environment as a reservoir for Carbapenem-resistant Arch Intern Med, 159: 1127-32, 1999.
organisms causing hospital-acquired infections - a systematic review 28 World Health Organization Global Priority List of Antibiotic-Resis-
of the literature. Clin Infect Dis, 64(10): 1435-1444, 2017. tant Bacteria to Guide Research, Discovery, and Development of
18 Garvey MI, Bradley CR, Bradley CW: Evaluating the risks of wash New Antibiotics. Geneva: WHO Press, p.1-7, 2017.

ERRATA CORRIGE

As an author of the article
"TWELVE YEARS OF LYELL'S SYNDROME IN THE BURN UNIT OF SÃO JOÃO HO-
SPITAL CENTER"
I declare that Pedro Reis, affiliated to the Department of Plastic Surgery and Burn Unit of
São João Hospital Center (Porto, Portugal) is a co-author and contributed for this work.

55

The new england journal of medicine

Original Article

Investigation of a Cluster of Sphingomonas
koreensis Infections

Ryan C. Johnson, Ph.D., Clay Deming, M.S., Sean Conlan, Ph.D.,
Caroline J. Zellmer, B.S., Angela V. Michelin, M.P.H., ShihQueen Lee‑Lin, M.S.,
Pamela J. Thomas, Ph.D., Morgan Park, Ph.D., Rebecca A. Weingarten, Ph.D.,

John Less, P.E., C.H.F.M., John P. Dekker, M.D., Ph.D.,
Karen M. Frank, M.D., Ph.D., Kimberlee A. Musser, Ph.D.,
John R. McQuiston, Ph.D., David K. Henderson, M.D., Anna F. Lau, Ph.D.,

Tara N. Palmore, M.D., and Julia A. Segre, Ph.D.​​

ABSTR ACT

BACKGROUND From the National Human Genome Re-
Plumbing systems are an infrequent but known reservoir for opportunistic micro- search Institute (R.C.J., C.D., S.C., S.L.-L.,
bial pathogens that can infect hospitalized patients. In 2016, a cluster of clinical J.A.S.), National Institutes of Health (NIH)
sphingomonas infections prompted an investigation. Clinical Center (C.J.Z., A.V.M., R.A.W.,
J.P.D., K.M.F., D.K.H., A.F.L., T.N.P.), and
METHODS the Division of Facilities, Operations, and
We performed whole-genome DNA sequencing on clinical isolates of multidrug- Maintenance (J.L.), NIH, Bethesda, and the
resistant Sphingomonas koreensis identified from 2006 through 2016 at the National NIH Intramural Sequencing Center, NIH,
Institutes of Health (NIH) Clinical Center. We cultured S. koreensis from the sinks Rockville (P.J.T., M.P.) — all in Maryland;
in patient rooms and performed both whole-genome and shotgun metagenomic Wadsworth Center, New York State De-
sequencing to identify a reservoir within the infrastructure of the hospital. These partment of Health, Albany (K.A.M.); and
isolates were compared with clinical and environmental S. koreensis isolates ob- the Special Bacteriology Reference Labora-
tained from other institutions. tory, Division of High-Consequence Patho-
gens and Pathology, Centers for Disease
RESULTS Control and Prevention, Atlanta (J.R.M.).
The investigation showed that two isolates of S. koreensis obtained from the six Dr. Park serves as an author on behalf of
patients identified in the 2016 cluster were unrelated, but four isolates shared the NIH Intramural Sequencing Center
more than 99.92% genetic similarity and were resistant to multiple antibiotic Comparative Sequencing Program. Ad-
agents. Retrospective analysis of banked clinical isolates of sphingomonas from dress reprint requests to Dr. Segre at the
the NIH Clinical Center revealed the intermittent recovery of a clonal strain over the National Human Genome Research Insti-
past decade. Unique single-nucleotide variants identified in strains of S. koreensis tute, NIH, Bldg. 49, Rm. 4A26, Bethesda,
elucidated the existence of a reservoir in the hospital plumbing. Clinical S. koreensis MD 20892, or at ­jsegre@­​nhgri​.­nih​.­gov;
isolates from other facilities were genetically distinct from the NIH isolates. Hos- or to Dr. Palmore at ­tpalmore@c​­ c​.n­ ih.​ g­ ov;
pital remediation strategies were guided by results of microbiologic culturing and or to Dr. Lau at ­anna​.­lau@­​nih​.­gov.
fine-scale genomic analyses.
Drs. Lau, Palmore, and Segre contributed
equally to this article.

N Engl J Med 2018;379:2529-39.
DOI: 10.1056/NEJMoa1803238

Copyright © 2018 Massachusetts Medical Society.

CONCLUSIONS
This genomic and epidemiologic investigation suggests that S. koreensis is an op-
portunistic human pathogen that both persisted in the NIH Clinical Center infra-
structure across time and space and caused health care–associated infections.
(Funded by the NIH Intramural Research Programs.)

n engl j med 379;26 nejm.org  December 27, 2018 2529

The New England Journal of Medicine
Downloaded from nejm.org on March 27, 2022. For personal use only. No other uses without permission.

Copyright © 2018 Massachusetts Medical Society. All rights reserved.

The new england journal of medicine

Health care–associated infections analyzed and cutoffs are applied to the number
affect 2 million patients each year in the of single-nucleotide variants (SNVs) that consti-
United States, and an increasing propor- tute a “match.”14 However, characterizing the
tion is attributed to multidrug-resistant bacte- underlying genetic diversity can provide critical
ria.1 Waterborne bacteria represent an important information to elucidate patient-to-patient trans-
subset of health care–associated pathogens and mission events or to identify reservoirs in the
are of increasing concern to public health authori- environment.
ties. In 2017, the Centers for Medicare and Med-
icaid Services established stringent requirements Although sphingomonas species are ubiqui-
for health care facilities to reduce the transmis- tous in natural and man-made aqueous environ-
sion risk of waterborne organisms from hospital ments, community-acquired infections are rarely
plumbing systems to patients.2 reported.15 The most common species implicated
as a human pathogen, S. paucimobilis, causes a
Historically, emphasis has been placed on pre- range of infections, including pneumonia, menin-
venting health care–associated transmission of gitis, catheter-associated bloodstream infections,
legionella and pseudomonas species; however, and wound infections.16,17
other waterborne organisms including stenotro-
phomonas, sphingomonas, burkholderia, and non­ Although the Centers for Disease Control and
tuberculous mycobacteria3-5 also pose risks for Prevention (CDC) urges that a single case of
health care–associated infections, particularly in health care–associated legionnaires’ disease should
immunocompromised patients. Exposure may trigger an epidemiologic investigation,18 inter-
occur through water droplets or water aerosols mittent infections with most other waterborne
that are inhaled or that breach normal defenses pathogens most likely go unrecognized, unless
through nonintact mucous membranes or inva- they are clustered temporally. Over a 6-month
sive devices.6-8 period in 2016, infections with sphingomonas
species developed in six inpatients at the Na-
Whole-genome sequence analyses have clari- tional Institutes of Health (NIH) Clinical Center.
fied epidemiologic chains of transmission for Isolates from four of these six patients were
outbreaks of Mycobacterium tuberculosis, carbapenem- identified as multidrug-resistant S. koreensis, a
resistant Klebsiella pneumoniae, methicillin-resistant nonfermenting gram-negative bacillus previous-
Staphylococcus aureus, and Legionella pneumophila.9-13 ly reported in only two clinical cases.19,20 This
Underlying genetic diversity can be a confounder cluster triggered an epidemiologic investigation
when single isolates from a given source are that used both culture-based and genomics-
based techniques to identify possible sources
and to inform effective intervention strategies.

8 Sphingomonas koreensis

7 Other sphingomonas species Methods

No. of Cases 6 New hospital Identification of Sphingomonas Cases
opened The epidemiologic investigation that was initiated
when the cluster of sphingomonas infections was
5 identified in 2016 included clinical and micro-
biologic record review and genomic analysis. Rec­
4 Hospital ords identified 37 patients from 2001 through
construction 2016 who had sphingomonas species cultured
from clinical specimens (Fig. 1; and Table S1 in
3 completed the Supplementary Appendix, available with the
full text of this article at NEJM.org). Environ-
2 mental and clinical S. koreensis isolates from
geographically distant institutions were solicited
1 for genomic and microbiologic comparisons. A
waiver of informed consent was granted by the
0 NIH Office of Human Subjects Research Protec-
tion for this study.
22222222222222222000000000000000001100010011000110154677189021233456

Figure 1. Sphingomonas Infections at the National Institutes of Health (NIH)
Clinical Center, According to Year.

A new NIH Clinical Center building was constructed in 2004 and opened
in 2005.

2530 n engl j med 379;26 nejm.org  December 27, 2018

The New England Journal of Medicine
Downloaded from nejm.org on March 27, 2022. For personal use only. No other uses without permission.

Copyright © 2018 Massachusetts Medical Society. All rights reserved.