0095-1137/06/$08.00⫹0 doi:10.1128/JCM.00547-06
Copyright © 2006, American Society for Microbiology. All Rights Reserved.
Comparison of
Acinetobacter baumannii
Isolates from the United
Kingdom and the United States That Were Associated with
Repatriated Casualties of the Iraq Conflict
Jane F. Turton,
1Mary E. Kaufmann,
1Martin J. Gill,
3Rachel Pike,
2Paul T. Scott,
4Joel Fishbain,
5David Craft,
5Gregory Deye,
6Scott Riddell,
6Luther E. Lindler,
4and Tyrone L. Pitt
1*
Laboratory of HealthCare Associated Infection1and Antibiotic Resistance Monitoring and Reference Laboratory,2Centre for Infections,
Health Protection Agency, London NW9 5EQ, United Kingdom; University Hospital Birmingham NHS Foundation Trust, Queen Elizabeth Medical Centre, Birmingham B15 2TJ, United Kingdom3; Walter Reed Army Institute of Research,
Silver Spring, Maryland 209104; Walter Reed Army Medical Center, Washington, D.C.5; and
Landstuhl Regional Medical Center, Landstuhl, Germany6
Received 14 March 2006/Returned for modification 22 April 2006/Accepted 12 May 2006
Acinetobacterisolates associated with casualties from the Iraq conflict from the United States were compared
with those from the United Kingdom by pulsed-field gel electrophoresis and integron analysis. Representatives of the main outbreak strain associated with casualties from both countries were indistinguishable in DNA profile. Two further outbreak strains were common to both sets of isolates.
Acinetobacter baumanniiis an increasingly important oppor-tunistic pathogen that affects vulnerable groups of people, par-ticularly patients in intensive care and burn units (5). It is usually associated with multiple antibiotic resistance, and few therapeutic agents remain that are effective against many strains of the organism (6, 20). AlthoughAcinetobacterspecies are commonly found in soil and water, the origin of the mul-tiresistant outbreak strains ofA. baumanniiin hospitals, which are often genetically similar, remains obscure (8, 14). Acineto-bacterinfections have frequently been reported in trauma vic-tims (1, 4, 7, 9, 12, 13, 15), but it is not clear whether these infections arose from environmental contamination at the scene or were acquired from the health care facilities in which these patients were subsequently treated.
Military and civilian casualties from the Iraq conflict have been repatriated to both the United Kingdom and the United States, and in both,Acinetobactercolonizations and infections have been reported. There has been some con-cern that multidrug-resistant outbreak strains ofA. bauman-niimay have been introduced into hospitals via these casu-alties. In the United Kingdom, some 10 strains, consisting of outbreak and sporadic strains, have been associated with casualties returning from Iraq, although 3 of the outbreak strains have previously been found in United Kingdom hos-pitals (Table 1). From the United States, there has been a published report of 102 patients withA. baumannii bactere-mias (4) and a media report (9a) of some 280 cases of infection in army medical centers that had received patients injured in the Iraq/Kuwait region or in Afghanistan. Some of these infections had been reported from Landstuhl Regional Medical Center, in Germany, which is a U.S. military regional medical center for
returning soldiers along the aeromedical evacuation route from Iraq/Kuwait to the United States and United Kingdom.
(A report of a United Kingdom nationwide survey of Acineto-bacter infections with possible links to Iraq, which includes some of the observations from this study, has been previously published [10].)
To identify if a common source(s) may be involved, 15 repre-sentative isolates ofA. baumanniifrom the outbreak investigation archive at Walter Reed Army Medical Center (WRAMC), a military tertiary care hospital in Washington, D.C., where many U.S. casualties from the war in Iraq receive care, were compared with United Kingdom isolates with links to Iraq received by the Laboratory of HealthCare Associated Infec-tion (Table 1). Comparison was by pulsed-field gel electrophore-sis (PFGE) of ApaI-digested genomic DNA, as described pre-viously (16). Class 1 integrase and blaOXA-23 genes were
detected using published methods (2, 11). Amplification and sequencing of the class 1 integron variable regions and deter-mination of the number of copies of orfX were carried out as previously described (17). Integrons were compared by cas-sette array size and HaeIII restriction digestion pattern. Iden-tification of outbreak strains asA. baumanniiwas by amplified rRNA gene restriction analysis (19).
A dendrogram of the PFGE profiles is given in Fig. 1. The United Kingdom isolates included representatives of the “T strain,” first seen in May 2003, which has been strongly asso-ciated with casualties from Iraq in hospital H1. They also included representatives of OXA-23 clone 2, a carbapenem-resistant clone withblaOXA-23, which was responsible for an
outbreak in hospital H2 in late 2003, following admission of a civilian casualty from Iraq. However, this clone had previously been found in three other United Kingdom hospitals with no known connection with Iraq. In addition, a comparatively mi-nor outbreak strain, associated with casualties from Iraq in two hospitals, was found. Profiles obtained of the U.S. isolates, collected between March and October 2003 from inpatients at
* Corresponding author. Mailing address. Laboratory of Health-Care Associated Infection, Centre for Infections, Health Protection Agency, 61, Colindale Avenue, London NW9 5EQ, United Kingdom. Phone: 44 (0)208 327 7224. Fax: 44(0)208 200 7449. E-mail: tyrone.pitt @hpa.org.uk.
2630
on May 16, 2020 by guest
http://jcm.asm.org/
WRAMC, Landstuhl Regional Medical Center, and a U.S. Army field hospital, Baghdad, Iraq, showed that these also included representatives of the T strain, OXA-23 clone 2, and the minor outbreak strain (referred to as H1AC-2, H3AC-1, or USAC-3) common to two hospitals in the United Kingdom. The remaining nine U.S. isolates, representing eight strains, were PCR negative for the class 1 integrase gene. Integrons are
[image:2.585.45.544.79.594.2]a characteristic of outbreak strains ofA. baumanniiand can be readily detected by integrase gene PCR; PCR-negative isolates are therefore likely to be sporadic (11, 17). These nine U.S. isolates represented different strains from those found in the United Kingdom. The United Kingdom isolates alone included representatives of the W strain, known to belong to European clone 1 (17). Isolates from both countries associated with Iraq
TABLE 1. United Kingdom isolates
Isolate
groupa Hospital
Date received (mo/yr)
PFGE result Link with Iraq Comments
Exchange isolates
UK 1 H1 5/03 T strain From a patient transferred from Iraq First isolate of T strain received by LHCAId; 1 of 5 isolates received (from 5
patients who served in Iraq) (UK 1 and 22–25), all representing different strains UK 2 H1 7/03 T strain Patient in hospital H1 Further representatives of the T strain from
patients in hospital H1 UK 3 H1 12/03 T strain Patient in hospital H1
UK 4 H1 11/04 T strain From soldier who served in Iraq immediately on arrival at the hospital
T strain linked directly with soldier returning from Iraq i.e., not acquired in hospital H1
UK 5 H1 12/04 T strain Patient in hospital H1
UK 6 H1 7/03 W strain
(European clone 1)
Patient in hospital H1 Strain predates Iraq conflictc
UK 7 H1 12/04 W strain
(European clone 1)
Patient in hospital H1 Strain predates Iraq conflictc
UK 8 H1 1/05 W strain
(European clone 1)
Patient in hospital H1 Strain predates Iraq conflictc
UK 9 H1 7/03 H1AC-3b Patient who served in Iraq PFGE profile similar to OXA-23 clone 2,
but OXA-23 negative by PCR UK 10 H1 9/04 H1AC-2 Patient who served in Iraq Represents same strain as H3AC-1 UK 11 H3 12/03 H3AC-1 Patient who served in Iraq Represents same strain as H1AC-2 UK 12 H2 10/03 OXA-23 clone 2 Outbreak in hospital H2 associated
with admission of civilian casualty from Baghdad
OXA-23 clone 2 previously found in 3 other UK hospitals, with the first isolate being seen in July 2003
UK 13 H2 10/03 OXA-23 clone 2 Patient in hospital H2 Further representative of OXA-23 clone 2 from outbreak in hospital H2
UK 14 H2 11/03 OXA-23 clone 2 Patient in hospital H2 Further representative of OXA-23 clone 2 from outbreak in hospital H2
UK 15 H1 7/03 Uniqueb Patient in hospital H1
UK 16 H4 8/03 Uniqueb Associated with Iraq
UK 17 H5 10/03 Unique Iraqi patient injured in Basra from gas cooker explosion
Treated in military hospital before coming to UK
UK 18 H6 11/03 Uniqueb Person who served in Iraq
UK 19 SE clone None Representative of widespread clone in UK
which has affected 40 hospitals between April 2000 and May 2005
UK 20 OXA-23 clone 1 None Representative of widespread clone in UK
that has affected 34 hospitals between November 2002 and May 2005
Additional isolates
UK 21 H7 4/03 Uniqueb Child with extensive burns from
Basra
UK 22 H1 5/03 V strain Patient who served in Iraq PFGE profile similar to, but distinct from, OXA-23 clone 2; OXA-23 negative by PCR
UK 23 H1 5/03 H1AC-2 Patient who served in Iraq Represents same strain as H3AC-1 UK 24 H1 5/03 R strain Patient who served in Iraq Strain previously found in hospital
UK 25 H1 5/03 W strain
(European clone 1)
Patient who served in Iraq Strain predates Iraq conflictc
a
Isolates UK 1 to UK 20 inclusive were included in an isolate exchange with WRAMC. b
This isolate was PCR negative for the class 1 integrase gene, indicating that it is sporadic (16). c
The strain was previously found in hospital H1; the clone has been widely found in Europe for decades (8). d
LHCAI, Laboratory of HealthCare Associated Infection.
on May 16, 2020 by guest
http://jcm.asm.org/
were distinct from representatives of the SE (southeast) clone (UK 19) and OXA-23 clone 1 (UK 20), which are the most prevalent strains ofA. baumannii in hospitals in the United Kingdom (16, 17).
To characterize these isolates further, the integrons in the United Kingdom and U.S. representatives of the outbreak strains were compared (Table 2). Most isolates of the T strain possessed a 3-kb integron. Isolate US 1, however, contained a 3.5-kb integron, which differed from the 3-kb integron (GenBank
accession number AY922990) in having an additional copy of orfX. U.S. representatives of the remaining outbreak strains carried the same integron as the corresponding United King-dom representatives. The integron cassette arrays in isolates of the T strain, SE clone, and OXA-23 clone 2 differ only in the number of copies of orfX and are associated with highly suc-cessful strains ofA. baumannii(17).
Antibiotic susceptibilities of representatives of the outbreak strains common to isolates from both countries are given in
FIG. 1. Dendrogram of PFGE profiles of ApaI-digested genomic DNA from isolates. United Kingdom isolates (UK 1 to UK 25) were from hospitals H1 to H7. Isolates from the United States (US 1 to US 15) were from anAcinetobacteroutbreak investigation archive at WRAMC.
on May 16, 2020 by guest
http://jcm.asm.org/
Table 3. Representatives of the T strain were resistant to ampicillin, amoxicillin-clavulanic acid, aztreonam, cefepime, cefotaxime, ceftazidime, cefoxitin, piperacillin, piperacillin-tazobactam, ciprofloxacin, gentamicin, and sulbactam; most United Kingdom isolates were susceptible to tobramycin and amikacin, but U.S. isolates were resistant. All the T strain isolates in this study were susceptible to carbapenems and colistin; however, some more recent isolates are carbapenem resistant (18). Isolates of OXA-23 clone 2 were resistant to all antibiotics tested, with the exception of amikacin, tobramycin (United Kingdom isolates only), minocycline, and colistin (all isolates). Isolates of the minor outbreak strain, referred to as H1AC-2, H3AC-1, or USAC-3, were also resistant to most antibiotics but susceptible to imipenem, meropenem, minocy-cline, sulbactam, and colistin.
The finding of strains that were common among the United Kingdom and U.S. isolates is consistent with a common origin. The fact that most isolates of the same strain also shared the same integron supports this. However, genetically similar iso-lates have often been found in geographically distant centers, perhaps because they have been independently selected from a common ancestor (8).
If a common origin for the outbreak strains is assumed, the source of these organisms is still a subject of some debate. Injured British and American soldiers had shared exposure to various medical facilities and modes of transport during their evacuation, and it is plausible that these may have been po-tential source(s). There are some claims that these infections may have originated from Iraqi soil, with a combination of acquisition from soil and selection by antibiotics having been
involved (9a). However,A. baumanniihas generally not been isolated from patients’ wounds immediately or very shortly after injury (21). Moreover, it seems unlikely that numerous individuals, injured at various sites, would all acquire the same outbreak strain in this way.
Of the three outbreak strains common to both the United Kingdom and U.S. isolates, it is the T strain that has been most strongly associated with these casualties and that has, in the United Kingdom at least, caused the most infections. The first known isolate of the T strain in the United King-dom was from a casualty returning from Iraq. It has been isolated from at least one soldier immediately on admission to the United Kingdom hospital, ruling out the possibility that it was acquired in that hospital. This strain has all the characteristics of a highly successful outbreak strain (and indeed shares many characteristics with the SE clone, which is widespread in London and southeast England). Although this strain has affected other patients in the hospital, it appears to have remained confined to that area for over 2 years and has therefore not greatly contributed to an in-crease in Acinetobacterinfections in the United Kingdom. Two isolates have, however, recently been found in a further United Kingdom hospital.
[image:4.585.40.544.81.324.2]In conclusion, at least one outbreak strain ofA. baumannii, responsible for further infections in the hospitals concerned, is associated with soldiers returning to the United States or United Kingdom from Iraq. The exact source(s) of the organ-ism remains unclear. The results of detailed epidemiological investigations conducted by the U.S. military will be reported separately.
TABLE 2. Class 1 integrons found in isolatesa
Isolate PFGE result
Approx size
(5⬘CSeto 3⬘CS) of
PCR product (kb)
Integron cassette array
United Kingdom and U.S. isolatesd
UK 1 T strain 3.0 aacC1orfX orfX orfX⬘aadA1ac
UK 2 T strain 3.0 aacC1orfX orfX orfX⬘aadA1ac
UK 3 T strain 3.0 aacC1orfX orfX orfX⬘aadA1ac
UK 4 T strain 3.0 aacC1orfX orfX orfX⬘aadA1ac
UK 5 T strain 3.0 aacC1orfX orfX orfX⬘aadA1ac
US 7 T strain 3.0 aacC1orfX orfX orfX⬘aadA1ac
US 9 T strain 3.0 aacC1orfX orfX orfX⬘aadA1ac
US 13 T strain 3.0 aacC1orfX orfX orfX⬘aadA1ac
US 1 T strain 3.5 aacC1orfX orfX orfX orfX⬘aadA1a
UK 14 OXA-23 clone 2 2.5 aacC1orfX orfX⬘aadA1a
US 12 OXA-23 clone 2 2.5 aacC1orfX orfX⬘aadA1a
UK 23 H1AC-2b 0.7 aadB
UK 10 H1AC-2b 0.7 aadB
UK 11 H3AC-1b 0.7 aadB
US 8 USAC-3b 0.7 aadB
Other United Kingdom isolates
UK 6 W strain 0.8 aacA4
UK 8 W strain 0.8 aacA4
UK 22 V strain 2.5 aacC1orfX orfX⬘aadA1a
UK 19 SE clone 3.0 aacC1orfX orfX orfX⬘aadA1a
UK 20 OXA-23 clone 1 2.3 aacA4 catB8 aadA1
a
UK 12, UK 13, and UK 17 were PCR positive for the class 1 integrase gene, but integron cassette arrays failed to amplify. b
Represents a single strain. c
As in GenBank accession no. AY922990. d
Comparison of integrons in United Kingdom and U.S. isolates of the same strain. e
5⬘CS and 3⬘CS are primers used to amplify integron cassette arrays (17).
on May 16, 2020 by guest
http://jcm.asm.org/
We are grateful to colleagues in United Kingdom hospital microbi-ology laboratories for sending the isolates in this study and to Derrick Crook for his collaboration. We are also grateful for technical support provided by Charla Gaddy and Tacita Hamilton at Walter Reed Army Institute of Research and Marie Pe and Kerry Snow at WRAMC.
The manuscript was reviewed by the Walter Reed Army Institute of Research. There is no objection to its publication. The opinions or assertions contained herein are the private views of the authors and are not to be construed as official or as reflecting true views of the De-partment of the Army or the DeDe-partment of Defense.
REFERENCES
1.Aarabi, B.1987. Comparative study of bacteriological contamination be-tween primary and secondary exploration of missile head wounds.
Neuro-surgery20:610–616.
2.Afzal-Shah, M., N. Woodford, and D. M. Livermore.2001. Characterization of OXA-25, OXA-26, and OXA-27, molecular class D beta-lactamases
as-sociated with carbapenem resistance in clinical isolates ofAcinetobacter
baumanniiAntimicrob. Agents Chemother.45:583–588.
3.Andrews, J. M.2005. BSAC standardized disc susceptibility testing method
(version 4). J. Antimicrob. Chemother.56:60–76.
4.Centers for Disease Control and Prevention.2004.Acinetobacter baumannii infections among patients at military medical facilities treating injured U.S.
service members, 2002–2004. Morb. Mortal. Wkly. Rep.53:1063–1066.
5.Chastre, J.2003. Infections due toAcinetobacter baumanniiin the ICU.
Semin. Respir. Crit. Care Med.24:69–78.
6.Coelho, J. M., N. Woodford, J. Turton, and D. M. Livermore.2004.
Mul-tiresistant acinetobacter in the UK: how big a threat? J. Hosp. Infect.58:
167–169.
7.Davis, K. A., K. A. Moran, C. K. McAllister, and P. J. Gray.2005. Multi-drug
resistantAcinetobacterextremity infections in soldiers. Emerg. Infect. Dis.
11:1218–1224.
8.Dijkshoorn, L., H. Aucken, P. Gerner-Smidt, P. Janssen, M. E. Kaufmann, J. Garaizar, J. Ursing, and T. L. Pitt.1996. Comparison of outbreak and
nonoutbreakAcinetobacter baumanniistrains by genotypic and phenotypic
methods. J. Clin. Microbiol.34:1519–1525.
9.Heath, C. H., C. T. Orell, R. C. E. Lee, J. W. Pearman, C. McCullough, and K. J. Christiansen.2003. A review of the Royal Perth Hospital Bali
experi-ence: an infection control perspective. Aust. Infect Control8:43–54.
9a.Herper, M.2 August 2005, posting date. The Iraq infection. Forbes http: //www.forbes.com/home/sciencesandmedicine/2005/08/02/iraq-war-infection -bacteria-cx_mh_0802iraqinfect.html.
10.Jones, A., D. Morgan, A. Walsh, J. Turton, D. Livermore, T. Pitt, A. Green, M. Gill, and D. Mortiboy.2006. Importation of multidrug-resis-tantAcinetobacterspp infections with casualties from Iraq. Lancet Infect.
Dis.6:317–318.
11.Koeleman, J. G. M., J. Stoof, M. W. van der Bijl, C. M. J. E. Vandenbroucke-Grauls, and P. H. M. Savelkoul.2001. Identification of epidemic strains of Acinetobacter baumanniiby integrase gene PCR. J. Clin. Microbiol.39:8–13. 12.Maegele, M., S. Gregor, E. Steinhausen, B. Bouillon, M. M. Heiss, W. Perbix, F. Wappler, D. Rixen, J. Geisen, B. Berger-Schreck, and R. Schwarz. 2005. The long-distance tertiary air transfer and care of tsunami victims: injury pattern and microbiological and psychological aspects. Crit. Care
Med.33:1136–1140.
13.Oncul, O., O. Keskin, H. V. Acar, Y. Kucukardali, R. Evrenkaya, E. M. Atasoyu, C. Top, S. Nalbant, S. Ozkan, G. Emekdas, S. Cavuslu, M. H. Us, A. Pasha, and M. Gokben.2002. Hospital-acquired infections following the
1999 Marmara earthquake. J. Hosp. Infect.51:47–51.
14.Rahal, J. J., and C. Urban.2000.Acinetobacter. Semin. Respir. Crit. Care
Med.21:341–348.
15.Tong, M. J.1972. Septic complications of war wounds. JAMA219:1044–1047. 16.Turton, J. F., M. E. Kaufmann, M. Warner, J. M. Coelho, L. Dijkshoorn, T. van der Reijden, and T. L. Pitt.2004. A prevalent, multiresistant, clone ofAcinetobacter baumanniiin southeast England. J. Hosp. Infect.58:170–179. 17.Turton, J. F., M. E. Kaufmann, J. Glover, J. M. Coelho, M. Warner, R. Pike,
and T. L. Pitt.2005. Detection and typing of integrons in epidemic strains of Acinetobacter baumanniifound in the United Kingdom. J. Clin. Microbiol.
43:3074–3082.
18.Turton, J. F., M. E. Ward, N. Woodford, M. E. Kaufmann, R. Pike, D. M. Livermore, and T. L. Pitt.2006. The role of ISAba1in expression of OXA
carbapenemase genes inAcinetobacter baumannii. FEMS Microbiol. Lett.
258:72–77.
19.Vaneechoutte, M., L. Dijkshoorn, I. Tjernberg, A. Elaichouni, P. de Vos, G. Claeys, and G. Verschraegen.1995. Identification ofAcinetobactergenomic species by amplified ribosomal DNA restriction analysis. J. Clin. Microbiol.
33:11–15.
20.Van Looveren, M., H. Goossens, and the ARPAC Steering Group.2004.
Antimicrobial resistance ofAcinetobacterspp. in Europe. Clin. Microbiol.
Infect.10:684–704.
21.Zapor, M. J., and K. A. Moran.2005. Infectious diseases during wartime.
Curr. Opin. Infect. Dis.18:395–399.
TABLE 3. MICs for representatives of outbreak strains of A. baumannii common to the United Kingdom and U.S. isolates Isolate Origin PFGE result MIC a(mg/liter) of: AMP (8) AMC (8) ATM (1) FEP (1) CTX (1) CAZ (2) FOX (8) PIP (16) TZP (16) IPM (4) MEM (4) CIP (1) COL (4) TOB (2) AMK (8) GEN (2) SUB (8) MIN (4 b) UK 12 United Kingdom OXA-23 clone 2 >64 >64 >64 32 >64 >64 >64 >64 >64 >32 16 >8.0 ⱕ 0.5 2 4 >32 32 0.5 UK 14 United Kingdom OXA-23 clone 2 >64 NT c NT 32 >64 >64 >64 >64 >64 8 16 >8.0 ⱕ 0.5 0.5 4 32 32 1 US 12 United States OXA-23 clone 2 >64 >64 >64 32 >256 >256 >64 >64 >64 16 16 >8.0 ⱕ 0.5 8 16 >32 32 1 UK 1 United Kingdom T strain >64 32 64 16 >256 64 >64 >64 >64 11 >8.0 ⱕ 0.5 0.5 2 >32 16 4 UK 3 United Kingdom T strain >64 >64 >64 16 >64 64 >64 >64 >64 12 >8.0 ⱕ 0.5 1 1 32 16 2 UK 4 United Kingdom T strain >64 32 64 8 256 64 >64 >64 64 11 >8.0 ⱕ 0.5 0.5 >64 16 16 2 UK 5 United Kingdom T strain >64 NT 64 16 >64 64 >64 >64 >64 12 >8 ⱕ 0.5 2 2 >32 32 2 US 1 United States T strain >64 32 64 16 >256 128 >64 >64 64 11 >8.0 ⱕ 0.5 1 1 >32 16 2 US 7 United States T strain >64 32 32 16 256 64 >64 >64 64 0.5 0.5 >8.0 ⱕ 0.5 16 >64 >32 16 1 US 9 United States T strain >64 64 64 16 >256 128 >64 >64 >64 11 >8.0 ⱕ 0.5 16 >64 >32 16 2 UK 10 United Kingdom H1AC-2 >64 >64 >64 32 >64 64 >64 >64 >64 22 >8.0 ⱕ 0.5 32 64 >32 4 0.5 US 8 United States USAC-3 >64 64 64 32 >256 64 >64 >64 >64 22 >8.0 ⱕ 0.5 >32 64 >32 41 aMICs were determined by agar dilution using British Society for Antimicrobial Chemotherapy (BSAC) methodology (3). Antibiotic abbreviations are a s follows: AMP, ampicillin; AMC, amoxicillin-clavulanic acid; ATM, aztreonam; FEP, cefepime; CTX, cefotaxime; CAZ, ceftazidime; FOX, cefoxitin; PIP, piperacillin; TZP, piperacillin-tazobactam; IPM, imipen em; MEM, meropenem; CIP, ciprofloxacin; TOB, tobramycin; AMK, amikacin; GEN, gentamicin; SUB, sulbactam; MIN, minocycline. Resistance is indicated by boldface. Breakpoints (mg/liter) are in parentheses foll owing the drug names. bNo BSAC breakpoint; Clinical and Laboratory Standards Institute breakpoint quoted. cNT, not tested.