J Clin Pathol: first published as 10.1136/jcp.33.3.288 on 1 March 1980. Downloaded from

J Clin Patho/ 1980; 33: 288-296

Antiseptic and resistance in Gram-negative bacteria causing urinary tract

DJ STICKLER AND B THOMAS

From the Department of Applied Biology, University of Wales Institute of Science and Technology. King Edward V1I A venue, Cardiff UK

SUMMARY A collection of 802 isolates of Gram-negative bacteria causing urinary tract was made from general practice, antenatal clinics, and local hospitals. The organisms were tested for their sensitivity to , cetrimide, glutaraldehyde, phenyl mercuric nitrate, a phenolic formulation, and a proprietary containing a mixture of picloxydine, octyl phenoxy polyethoxyethanol, and benzalkonium chloride. Escherichia coli, the major species isolated, proved to be uniformly sensitive to these agents. Approximately 10°,! of the total number of isolates, however, exhibited a degree of resistance to the cationic agents. These resistant organisms were members of the genera Proteus, Providencia, and Pseudomonas; they were also generally resistant to five, six, or seven . It is proposed therefore that an antiseptic policy which involves the intensive use of cationic might lead to the selection of a flora of notoriously drug- resistant species. copyright.

During the course of studies on the development of minimum inhibitory concentrations (MICs) of up urinary tract infection in patients undergoing to 800 ,sg/ml chlorhexidine,'4 well above the level of intermittent bladder catheterisation in the early 10-50 jug/ml orginally reported to inhibit the growth stages of paraplegia, observations were made on the of Gram-negative species.5 effect of the repeated application of the antiseptic These chlorhexidine-resistant strains were isolated http://jcp.bmj.com/ chlorhexidine on the bacterial flora of the urethral from an unusual clinical situation in which repeated meatus.' 2 The bladder management of these patients exposure to the antiseptic over periods of up to involved the washing of the penis with aqueous 3-4 weeks takes place. It is perhaps understandable solutions of chlorhexidine (600 jtg/ml) before the that resistant strains should be isolated from a insertion of the catheter at approximately 8-hour situation where such a strong selective pressure is intervals. The flora was examined daily before and operating. However, as resistance to agents like after application of the antiseptic, and the general chlorhexidine is not generally screened for in clinical on September 24, 2021 by guest. Protected pattern that emerged from this work was that for laboratories it is not possible to say whether resis- the first few days after trauma the meatus carried a tance to antiseptics is limited to such special cir- Gram-positive flora which the chlorhexidine usually cumstances or is a more general phenomenon. The eliminated. A Gram-negative population developed objective of this study was to observe the sensitivity subsequently, however, and proved to be more of Gram-negative organisms causing urinary tract refractory to chlorhexidine. In particular, Proteus infections, in a variety of types of patients, to a mirabilis, Pseudomonas aeruginosa, Providencia number of antiseptics and disinfectants. A col- stuartii, and Klebsiella frequently survived the lection of organisms was therefore made from application of the antiseptic, and in laboratory general practice, antenatal clinics, and inpatients at tests many of the strains demonstrated an ability to five local hospitals. The sensitivity of these isolates grow in media containing 200 ,ug/ml of the agent. to chlorhexidine, cetrimide, glutaraldehyde, phenyl Subsequently, some of the Pr. mirabilis and P. mercuric nitrate, a phenolic formulation, and a stuartii strains from this source were shown to have proprietary antiseptic containing a mixture of picloxydine, octylphenoxy polyethoxyethanol, and benzalkonium chloride was determined. In addition, Received for publication 5 September 1979 the antibiotic resistance patterns were recorded to 288 J Clin Pathol: first published as 10.1136/jcp.33.3.288 on 1 March 1980. Downloaded from

Antiseptic and antibiotic resistance in Gram-negative bacteria causing urinary tract infections 289 explore any possible relationship between disinfect- colony formation after overnight incubation at 370C ant and drug resistance. was taken as the MIC for that strain. The tests were routinely performed in duplicate and, with each Methods and material batch of strains tested, E. coli 10418 and Pr. mirabilis 61 were included as control reference organisms with BACTERIAL STRAINS known MICs. The clinical isolates used in this study all originated from cases ofsignificant bacilluria ( > 105 viable cells/ ANTIBIOTIC SENSITIVITY ml urine). They were obtained over a 12-month The antibiotic sensitivity patterns of the isolates were period from a hospital laboratory which served determined by seeding plates of DST agar (Oxoid several hospitals and local health centres in the Ltd) with young log phase broth cultures (0-1 ml) Cardiff area. To avoid duplication, repeated and applying U3 Multodiscs (Oxoid Ltd), which are isolates of the same species from individual patients impregnated with the following antibiotics: gen- were excluded from the survey. The organisms were tamicin (10 )ug), colistin (10 ,pg), nitrofurantoin identified using the methods of Cowan and Steel6 (200 pug), sulphafurazole (500 tig), kanamycin (30 and in some cases with the API 20E Microtube ,ug), (25 jug), sulphamethoxazole/tri- System (API Products Ltd). Escherichia coli NCTC methoprim (25 ,tg), and (50 ,ug). E. 10418 and Pr. mirabilis 613 were used as control coli NCTC 10418 was used as sensitive control with organisms in the sensitivity testing. each batch of strains tests.

ANTIBACTERIAL AGENTS Results The chlorhexidine digluconate and the cetrimide were obtained from ICI Ltd. Glutaraldehyde was In a 12-month period, over 1000 isolates of Gram- purchased from Sigma Ltd and phenyl mercuric negative bacteria were obtained from urinary nitrate from BDH Ltd. The phenolic preparation infections. E. coli proved to be by far the most

contained 16% v/v of a mixture of chloroxylenol, prevalent species, and after six months when some copyright. chlorocresol, chlorophene, sodium 0-phenylphenate, 369 E. coli isolates had been investigated, the sen- and sodium pentachlorophenate in a basis of sitivity testing of this species was discontinued. The anionic emulsifier and alcohol (Hycolin, Pearson species distribution of the isolates examined in the Ltd. Clough Road, Hull). The sixth antibacterial survey is shown in Table 1. The results of the sen- agent was the proprietary antiseptic Resiguard sitivities of these organisms to the six antibacterials (Nicholas Laboratories Ltd) which contains pic- are summarised in Figures 1 to 6. The histograms loxydine digluconate (1 % w/v), octyl phenoxy represent the percentage of isolates having each MIC, http://jcp.bmj.com/ polyethoxyethanol (11-0% w/v), and benzalkonium grouped on the basis of the multiplicity of their chloride (12% w/v). antibiotic resistance. It can be seen that E. coli was uniformly sensitive to the antibacterial, having MIC DETERMINATIONS MICs of the six agents equivalent to that of the As the inoculum history of bacterial cultures used in control E. coli strain 10418 and well below the antimicrobial sensitivity tests may affect the results recommended use concentrations of these agents.

obtained7 a standard procedure was adopted for The situation is different with Proteus, P. stuartii, on September 24, 2021 by guest. Protected testing the clinical isolates. A single colony from a isolated urinary tract pure culture of each strain growing on MacConkey Table 1 Bacterial species from infections over a 12-month period agar after overnight incubation at 370C was inocu- lated into nutrient broth (Oxoid Nutrient Broth No. Species Number of isolates 2). After 18 hours' incubation at 370C, volumes (5 of a 10-4 dilution of the broth culture, each E. coli 369* Atl) Pr. mirabilis 139 containing approximately 103 viable cells, were Pr. morganii 31 dropped on to the surface of nutrient agar plates Pr. vulgaris 10 Pr. rettgeri 1 containing various concentrations of the anti- Klebsiella 167 bacterials. A standard method was adopted for the P. stuartii 24 The were added to Ps. aeruginosa 35 preparation of the plates. agents Citrobacter 12 molten agar that had been allowed to cool to 500C, Enterobacter 9 and the plates were poured immediately. When set, Acinetobacter 3 then Serratia marcesens 2 the plates were dried for 20 minutes at 370C and Total 802 used directly. The lowest concentration of the agent preventing *Number of isolates obtained in first six months of the study. J Clin Pathol: first published as 10.1136/jcp.33.3.288 on 1 March 1980. Downloaded from 290 Stickler and Thomas No of isolates in each group MIC of Multiplicity E col' 10418 of antibiotic <.1fA/m resistance

100

u

U

0 0) 52 ,u

0 0C

29 copyright.

E col/i Proteus Klebsiel /a

Pseudomonas http://jcp.bmj.com/ Providen cia I Other species 5 a 10 50 200 800 1600 ;10 50 200 800 1600 MIC Chlorhexidine (Jug/ml) on September 24, 2021 by guest. Protected Fig. 1 Sensitivity of the isolates to chlorhexidine.

and Pseudomonas, where MICs substantially higher antiseptics and disinfectants are concerned. For the than those of the control strain were recorded for the purpose of this study it was decided that if the MIC cationic antiseptics. value for an organism was greater than the con- In deciding whether strains are resistant or centration of the agent normally recommended for sensitive to antiseptics or disinfectants, it is difficult routine use, then that isolate was designated as to know what criterion to apply. With antibiotics it resistant. When this criterion is applied (Table 2) it is usual to designate organisms as resistant if they appears that all the isolates were sensitive to are able to multiply in the maximum concentration glutaraldehyde, the phenolic mixture, and phenyl- of the drug attainable at the site of the infection. mercuric nitrate. A small percentage of the total were This sort of yardstick is clearly not applicable where resistant to the cationic agents. J Clin Pathol: first published as 10.1136/jcp.33.3.288 on 1 March 1980. Downloaded from

Antiseptic and antibiotic resistance in Gram-negative bacteria causing urinary tract infection 291

4

u

5-) u 5 C-

tn a)

0 6 copyright. 7

<10 50 200 800 160( http://jcp.bmj.com/ MIC (jug/ ml ) Cel rimide

Fig. 2 Sensitivity of the isolates to cetrimide on September 24, 2021 by guest. Protected

Table 2 Percentage ofisolates ofeach species with MIC values greater than the normal recommended use concentration of each antibacterial

Organism Chlorhexidine Cetrimide Picloxydine mixture Glutaraldehyde Phenolic mixture Phenyl mercuric MIC > 500 jsgIml MIC > 1000 ,g/ml MIC > 0625Y. v/v MIC > 2% v/v MIC > 2%, v/v nitrate MIC > 20iMg/ml E. coli 00 00 00 00 00 00 Klebsiella 1-8 0 0 0 9 0.0 0.0 0 0 Proteus 38-7 2-2 35 9 0 0 0.0 0.0 P. stuartii 83-3 29-2 75 0 0.0 0.0 0.0 Ps. aeruginosa 45-7 65-7 60-0 0.0 0 0 0 0 Others 3-8 3-8 7-6 0.0 0.0 0.0 % of total No. 13-7 4-4 - 13-0 0 0 0.0 0 0 of isolates J Clin Pathol: first published as 10.1136/jcp.33.3.288 on 1 March 1980. Downloaded from

292 Stickler and Thomas

M IC of Multiplicity E coli 10418 of antibiotic resistance < 0020/o0 50 - 0

& II.

t6 50 - P-414-A ur au 2.1:;=

2 Ln 50 0 #--

50 copyright. 3 http://jcp.bmj.com/ ~002 01 0-2 0-5 1-0 2-0 -002 01 02 0-5 10 20 MIC Picloxydine concentrated v/v)

Fig. 3 Sensitivity of the isolates to the picloxydine preparation. on September 24, 2021 by guest. Protected

A comparison between the chlorhexidine sen- that have been reported to be responsible for acute, sitivity of isolates from hospital inpatients and those chronic, and hospital-acquired urinary infection ;8 910 from general practice and antenatal clinics is shown the most common organism, E. coli, proved to be in Figure 7. It is clear that resistant organisms are extremely sensitive to all the antibacterial, the MIC mainly found in infections originating in hospital. values recorded for the isolates being equivalent to that of the control culture collection strain E. coli Discussion 10418 and well below the normally recommended use concentrations for the six agents. The variety of Gram-negative species isolated and The criterion used to describe strains as resistant tested in this survey reflects the range of organisms was an MIC greater than the normally used con- J Clin Pathol: first published as 10.1136/jcp.33.3.288 on 1 March 1980. Downloaded from

Antiseptic and antibiotic resistance in Gram-negative bacteria causing urinary tract infection 293

MIC of Multiplicity of E c ol antibiotic 10418`- 0./ resistance 0 50 4

C) rrv-- VLA l -ia 5 -uU 0 50 a) .

U) 0

0en 0 6 50- 00

7I 50 copyright.

<001 002 004 008 0 16 032 001 002 004 008 016 032 MIC Phenolic concentrate (0/.v/v) http://jcp.bmj.com/

Fig. 4 Sensitivity of the isolates to the phenolic agent. on September 24, 2021 by guest. Protected

centration of the agent. Although it was fully was limited to the cationic agents and to the genera realised that the MIC value for a disinfectant may Proteus, Providencia, and Pseudomonas. There have not represent the concentration that will be effective been previous reports of this phenomenon. Chlor- in practice, some support for the relevance of such a hexidine resistance has been reported in a Pseudo- standard was provided by Thomas et al.11 who monas sp. 12, Ps. aeruginosa,'3 Pr. mirabilis,314 and showed that urine-grown cultures of organisms with P. stuartii4 In all these instances, the organisms were MICs of chlorhexidine > 800 ,tg/ml survived well isolated from sources where they had been exposed when challenged with the recommended level of this to the antiseptic. In the present survey, Fig. 7 antiseptic. When the criterion was applied to the shows that the resistant strains came predominantly results of this survey it can be seen that the resistance from inpatient infections. In fact, the majority came J Clin Pathol: first published as 10.1136/jcp.33.3.288 on 1 March 1980. Downloaded from

294 Stickler and Thomas MIC of Multiplicity E coli 10418 of antibiotic 0.08°/. resistance 4

U - 5

up 4)

Ul 4) 4- 03: 00 6 0 copyright.

7 http://jcp.bmj.com/

e002 004 008 016 MIC Glutaraldehyde (/. v/v)

5

Fig. Sensitivity of the isolates to glutaraldehyde. on September 24, 2021 by guest. Protected

from chronic infections in geriatric or paraplegic The general conclusion from the survey is that patients. Infections in these cases are often with resistance to antiseptics and disinfectants is not a mixed populations, and the observation that the widespread phenomenon in organisms that are antiseptic-resistant organisms were also generally responsible for urinary tract infections. Resistance resistant to five, six, or seven antibiotics (Figs 1-3) appears to be limited to the genera Pseudomonas, leads to speculation that the challenge of such a Proteus, and Providencia and to the cationic anti- mixed culture with, for example, an instillation of septics. The resistant strains in this survey were chlorhexidine into the bladder might result in the mostly isolated from patients in whom chlorhexidine selection of the chlorhexidine-resistant, drug- was being used to wash down the urethral meatus resistant species. before insertion of a bladder catheter. A chlorhexi- J Clin Pathol: first published as 10.1136/jcp.33.3.288 on 1 March 1980. Downloaded from

Antiseptic and antibiotic resistance in Gram-negative bacteria causing urinary tract infection 295 MIC of E.coli 10418

02 ug /mln

-c C)u -c

LA i3:

0 4-

0o copyright. http://jcp.bmj.com/

<0-1 0-2 0-5 1-0 2-0 40 <01 0-2 0-5 10 20 40 MIC PMN (ug/ml)

Fig. 6 Sensitivity of the isolates to phenyl mercuric nitrate (PMN). on September 24, 2021 by guest. Protected

dine-containing cream was used to facilitate the The recent report'5 that resistance to the anti- passage of the catheter through the uretha, and in septic hexachlorophane in Ps. aeruginosa was some cases the antiseptic was incorporated into the determined by an extrachromosomal element which urine drainage bag. These observations, together also carried the genes for sulphonamide and gen- with an examination of the antibiotic resistance tamicin resistance has important implications and patterns of the isolates, have led us to the con- has led us to consider whether the observed cor- clusion that an antiseptic policy which relies on the relation between antibiotic and cationic antiseptic intensive use of cationic antiseptics such as chlor- resistance has a basis in an association of the hexidine could well result in the selection of a resistance genes. The possibility of such a relation- flora of notoriously drug-resistant species. ship existing between the genetic factors for chlor- J Clin Pathol: first published as 10.1136/jcp.33.3.288 on 1 March 1980. Downloaded from

296 Stickler and Thomas

100 - 5 Davies GE, Francis J, Martin AR, Rose FL, Swain G. In - patients 760/1 ot total 1 :6-di-4'-chlorophenyldiguanidohexane (Hibitane): laboratory investigation of a new antibacterial agent of high potency. Br J Pharmacol Chemother 1954;9: 192-6. a50- 6 Cowan ST, ed. Cowan and Steel's Manual for the _c Identification of Medical Bacteria. 2nd edition. 50 London: Cambridge University Press, 1974. Farewell JA, Brown MRW. The influence of inoculum a history on the response of microorganisms to inhibitory and destructive agents. In: Hugo WB, ed. Inhibition and Destruction of the Microbial Cell. l100 Out- patients 24°1. of total London: Academic Press, 1971:703-52. 8 Mond NC, Percival A, Williams JD, Brumfitt W. Presentation, diagnosis and treatment of urinary- so tract infections in general practice. Lancet 1965; 1:514-6. 9 Milner PF. The differentiation of Enterobacteriaceae infecting the urinary tract: a study in male para- plegics. J Clin Pathol 1963;16:39-45. 10 Brumfitt W. Penicillin and related antibiotics: Gram- negative infections: urine. J R Coil Physicians Lond <10 50 200 800 1600 1972;6:194-202. MIC .Chlorhexidine (,ug/ml) Thomas B, Sykes L, Stickler DJ. Sensitivity of urine- Fig. 7 A comparison ofthe chlorhexidine sensitivity grown cells of Providencia stuartii to antiseptics. J ofisolates from inpatients and Clin Pathol 1978 ;31:929-32. outpatients. 12 Bentley M, Davis A, Field BS, Roberts W. Character- istics of growth of a Pseudomonas species in a hexidine and antibiotic resistance in P. stuartii is now culture media containing chlorhexidine (Abstract). copyright. under investigation in this laboratory. Biochem J 1968;110:46p. 13 Brown MRW, Watkins WM, Scott Foster JH. Step- wise resistance to polymyxin and other agents by References Pseudomonas aeruginosa (Abstract). J Gen Microbiol 1969 ;55 :xvii-xviii. Stickler DJ, Wilmot CB, O'Flynn JD. The mode of 14 Gillespie WA, Lennon GG, Linton KB, Phippen GA. development of urinary infection in intermittently Prevention of urinary infection by means of closed http://jcp.bmj.com/ catheterized male paraplegics. Paraplegia 1971; drainage into a sterile plastic bag. Br MedJ 1967;3: 8:243-52. 90-2. 2 O'Flynn JD, Stickler DJ. Disinfectants and Gram- 15 Sutton L, Jacoby GA. Plasmid-determined resistance negative bacteria (Letter). Lancet 1972;1 :489-90. to hexachlorophene in Pseudomonas aeruginosa. 3 Stickler DJ. Chlorhexidine resistance in Proteus Antimicrob Agents Chemother 1978;13:634-6. mirabilis. J Clin Pathol 1974;27:284-7. 4 Stickler DJ, Thomas B. Sensitivity of Providence to Requests for reprints to: Dr DJ Stickler, Department of antiseptics and disinfectants. J Clin Pathol 1976;29: Applied Biology, Uwist, King Edward VII Avenue, on September 24, 2021 by guest. Protected 815-23. Cardiff CF1 3NU, UK.