Microbial contamination of and in four major hospitals in Trinidad

Tswana Gajadhar,1Alicia Lara,1Patricia Sealy,1 and Abiodun A. Adesiyun2

ABSTRACT Objective. To assess the microbial contamination of disinfectants and antiseptics in major hospitals on the Caribbean island of Trinidad. Methods. For this cross-sectional study, disinfectants and antiseptics were sampled from the pharmacy departments, the pediatric/neonatal wards, and the surgical wards of four hos- pitals. The samples were cultured for aerobic on nutrient agar using the surface plat- ing method. The sensitivity of bacterial isolates was determined by the disk diffusion method, using 14 agents. We studied a total of 180 / sam- ples: 60 of gluconate (Hibitane), 60 of chlorhexidine gluconate and (Savlon), and 60 of methylated spirit. Results. Of the 180 samples studied, 11 of them (6.1%) were contaminated by aerobic bac- teria. All bacteria isolated were Pseudomonasspp. Of the 11 contaminated samples, 6 of them (54.5%) occurred at the pharmacy level while 5 (45.5%) were from diluted pre-use or in-use samples in the pediatric/neonatal wards or the surgical wards. Chlorhexidine gluconate and accounted for 9 of the 11 contaminated disinfectants/antiseptics (81.8%), and chlorhexidine gluconate accounted for the remaining 2 (18.2%). Only two of the four hospitals had contaminated disinfectant/antiseptic samples. All 24 isolates of Pseudomonasspp. tested were resistant to one or more of the 14 antimicrobial agents tested, with the prevalence of resistance to ciprofloxacin, norfloxacin, tobramycin, and gentamicin being 58.3%, 50.0%, 45.8%, and 41.7%, respectively. Conclusions. Our results show that contaminated disinfectants/antiseptics pose a health risk to patients, particularly in the pediatric and surgical wards. The high prevalence of resis- tance to antimicrobial agents exhibited by the Pseudomonasspp. that were isolated is of spe- cial therapeutic concern.

Key words Disinfectants; anti-infective agents, local; hospitals, control; Pseudomonas; Trinidad and Tobago.

1 The University of the West Indies, Faculty of Med- Disinfectants and antiseptics of dif- ment of new preparations as well as ical Sciences, Pharmacy Programme, St. Augus- ferent preparations and concentrations the introduction of various strategies tine, Trinidad and Tobago. 2 The University of the West Indies, Faculty of Med- have been used universally on ani- to achieve optimal efficacy and effec- ical Sciences, School of Veterinary Medicine, St. mate and inanimate surfaces in hospi- tiveness of disinfectants and antisep- Augustine, Trinidad and Tobago. Send correspon- dence to: Professor A. A. Adesiyun, School of Vet- tals (1–4). The observed increase in the tics (5, 6). erinary Medicine, Faculty of Medical Sciences, resistance of a number of bacterial It has been documented that stock University of the West Indies, St. Augustine, Trinidad and Tobago; e-mail: abiodunadesiyun@ pathogens to some disinfectants and solutions of disinfectants/antiseptics hotmail.com antiseptics has resulted in the develop- sold commercially, diluted disinfec-

Rev Panam Salud Publica/Pan Am J Public Health14(3), 2003 193 tants/antiseptics prepared in hospital and other health institutions) and pri- located in the northern region of pharmacies for distribution to the var- vate establishments (hospitals, clinics, Trinidad, and the fourth is in the ious hospital departments, and disin- and nursing homes). The reform is an southern part of the island. fectants/antiseptics that are in use in attempt to improve the delivery of hospital departments have been ex- care to its citizens, in- posed to contamination (7, 8). The lev- cluding reducing or eliminating conta- Laboratory study of samples of els of contamination by some of these mination of disinfectants/antiseptics disinfectants collected in 4 hospitals bacteria have ranged from 102to 108 as well as of surgical equipment in the colony-forming units per mL of disin- local hospitals. This cross-sectional study was de- fectant/antiseptic, with the possibility signed to determine the bacteriological of attaining an infective dose at the site quality of disinfectants/antiseptics of application (8, 9). The public health MATERIALS AND METHODS used in the four hospitals that we had significance of the use of contaminated selected. Each of the four hospitals disinfectants/antiseptics in neonatal This study consisted of two parts: was visited a total of four times, usu- units, surgical wards, and intensive (1) a questionnaire component, used ally a week apart. Samples were col- care units cannot be overemphasized. with 11 hospitals, and (2) a laboratory lected at the three levels mentioned Reports of epidemics of nosocomial study on samples of disinfectants/ earlier: pharmacy departments, pedi- bacterial have implicated antiseptics collected in 4 of those 11 atric/neonatal wards, and surgical contaminated disinfectants/antisep- hospitals. wards. The samples from each of the tics both applied directly to the skin of three levels at the four hospitals were animals or humans and used to decon- given code numbers in order to avoid taminate instruments and appliances Identification of commonly used any bias in their analysis. used for diagnosis and treatment (10, disinfectants and antiseptics The study period lasted from Janu- 11). It has also been reported that con- in 11 hospitals ary 2002 to March 2002. During each taminated disinfectants exhibit de- visit to the four hospitals, in the phar- creased efficacy and effectiveness (12, Initially, 11 health institutions in macy departments we collected and 13). Also of therapeutic significance Trinidad were surveyed using a pre- placed approximately 10 mL each of are reports that a number of bacterial tested standardized questionnaire to the pharmacy stock and of the diluted contaminants isolated from disinfec- elicit information on disinfectant/anti- pre-pediatric/pre-neonatal and pre- tants have exhibited resistance to com- septic use. The questionnaire, which samples into sterile universal monly used antimicrobial agents (14). was administered to the head nurse bottles for our analysis. The pedi- In the country of Trinidad and To- and pharmacist, asked about the fol- atric/neonatal wards and the surgical bago there is a dearth of information lowing subjects: types of disinfec- wards served as the sources of diluted on the microbial contamination of dis- tants/antiseptics used, methods of pre-use and in-use samples. All the infectants/antiseptics used in health preparation of disinfectants/antisep- samples were transported ice-cooled institutions. This study was therefore tics, conditions under which disin- to the Public Health Laboratory of the conducted to determine the preva- fectants/antiseptics are prepared, School of Veterinary Medicine of the lence of bacteria in disinfectants/anti- identification of potential sources of University of the West Indies, which is septics at three levels: (1) pharmacy contamination, staff responsible for located in Champ Fleurs, Trinidad. departments (pharmacy stock and preparation and dilution of disinfec- The samples arrived at the laboratory diluted samples before delivery to tants/antiseptics, and potential use of within 2 h of collection and were wards), (2) pediatric/neonatal wards disinfectants/antiseptics in the hospi- processed immediately upon arrival. (after delivery from the pharmacy de- tals. Due to constraints with time, lo- At the laboratory, 1 mL of each of partment, diluted pre-use and in-use gistics, and resources, the main hospi- the samples was added to 9 mL of nu- samples), and (3) surgical wards (after tal in Tobago, the sister island, was not trient broth in new sterile universal delivery from the pharmacy depart- included in this study. bottles. This 1:10 dilution of the sample ment, diluted pre-use and in-use sam- After analyzing the data from the was thoroughly mixed using a vortex ples). The investigation also deter- questionnaires, we selected four mixer. To quantify the aerobic bacteria mined the sensitivity of the bacteria to health institutions, as these four pre- in the samples, we used the procedure antimicrobial agents. pared and utilized the largest volumes described by Khor and Jegathesan (15). Trinidad and Tobago, a twin-island of disinfectants and antiseptics pur- With a micropipette, the 1:10 dilution republic in the Caribbean, has a popu- chased by the Government and dis- of sample in nutrient broth was inocu- lation of 1.3 million people. Its health tributed to the pharmacy departments lated onto a dry nutrient agar plate. care system, which is currently under- of these public health care institutions. For each nutrient agar plate a total of going structural reform, consists of a These four hospitals are the largest 10 equally spaced 20-mL inocula were combination of public establishments public health care institutions in placed on the plate. Thus, a total of (government hospitals, health centers, Trinidad. Three of these hospitals are 200 mL of the 1:10 diluted sample

194 Gajadhar et al. • Microbial contamination of disinfectants and antiseptics in four major hospitals in Trinidad was added to each plate. For each sam- zole/trimethoprim (30 mg), tal attendants attached to the pharmacy ple, duplicate inoculated plates were (10 mg), nalidixic acid (30 mg), chlo- departments performed this role in the prepared. The inoculated plates were ramphenicol (30 mg), ciprofloxacin other 4 hospitals. allowed to dry at room temperature (5 mg), norfloxacin (10 mg), to- All the disinfectants/antiseptics (25oC). Subsequently, one inoculated bramycin (10 mg), gentamicin (10 mg), were distributed from the pharmacy plate was incubated at 25 oC for 72 h, kanamycin (30 mg), and departments to the pediatric/neonatal and the other plate was incubated at (30 mg). Each isolate was incubated at wards and the surgical wards in recy- 37 oC for 24 h, both under aerobic con- the temperature and for the duration cled, nonsterile, multiuse containers ditions. The nutrient agar plates were used for primary isolation, i.e., 25 °C provided by the pharmacy depart- observed for bacterial growths. The for 72 h or 37 °C for 24 h. ments. In some cases the containers total count of colonies in the 10 areas The prevalences of bacterial contam- were recycled bottles that originally (i.e., 200 mL) was determined and mul- ination and of antibiotic resistance of contained the concentrated disinfec- tiplied by 5 to determine the number of isolates were compared for all sam- tants/antiseptics, and in other in- colonies per mL (1 000 mL) of sample. ples using the chi-square (x2) test with stances they were glass bottles bought Since each sample was initially diluted Epi Info version 6.02 computer soft- for this purpose. Concerning the use 1:10 in nutrient broth prior to plating ware (Centers for Disease Control and of antiseptics to clean skin surfaces, on the nutrient agar plates, the number Prevention, Atlanta, Georgia, United chlorhexidine gluconate (Hibitane) of colony-forming units per mL of States of America). and chlorhexidine gluconate and cetri- disinfectant/antiseptic was determined monium bromide (Savlon) were most by a further multiplication of the count frequently used, each reportedly used by 10. RESULTS by 6 of the 11 health facilities (54.5%). To qualitatively determine the pres- Prior to administering an injection, 2 ence of aerobic bacteria in the samples, Questionnaire survey of the 11 hospitals (18.2%) used swabs new duplicate sterile universal bottles of 11 hospitals containing methylated spirit and 70% were used. One of the bottles contain- to clean the skin. To clean ing inoculated nutrient broth (1:10) According to the surveys done with inanimate surfaces, chlorhexidine glu- was incubated at 25 °C for 72 h, and the 11 hospitals, some of them used conate was the most frequently used the other bottle was incubated at 37 oC more than one source of water for dilut- substance, used by 5 of the 11 hospitals for 24 h, both under aerobic condi- ing different disinfectants/antiseptics (45.5%). Chlorhexidine gluconate plus tions. Growth in each broth was inocu- and for washing containers. Three of cetrimonium bromide was used by 3 lated, without dilution, onto nutrient the health facilities used distilled wa- of the 11 hospitals (27.3%) to clean agar plates and incubated under the ter as diluent for preparing diluted wounds. conditions to which the broths were disinfectants/antiseptics, 1 used boiled initially exposed, that is, at 25 °C for water, and 9 used pipeborne water 72 h and at 37 oC for 24 h, respectively. supplied by the Water and Sewerage Laboratory study of samples from The outcome was expressed as posi- Authority (WASA), which is the Gov- 4 hospitals tive or negative for aerobic bacteria. ernment agency responsible for the Colonies representative of the vari- treatment and distribution of potable From the 11 health facilities that we ous growths detected on the nutrient pipeborne water throughout Trinidad surveyed, 4 of them were selected to agar plates that had been incubated at and Tobago. WASA water was directly carry out laboratory studies to deter- 25°C and at 37°C were picked and used at 7 of the hospitals, while the mine the bacteriological quality of the streaked for isolation on blood agar other 2 used WASA water that had disinfectants/antiseptics that they uti- plates and incubated at 25 °C for 72 h been stored in large tanks within the lized. We chose the 4 hospitals because and at 37 °C for 24 h, respectively. All hospital premises. None of the 11 in- they used the greatest volumes of dis- bacterial isolates were identified using stitutions used sterilized utensils in infectants/antiseptics (chlorhexidine standard methods (16). Confirmation preparing the disinfectants/antisep- gluconate, chlorhexidine and cetrimo- of the isolates was done at the Carib- tics, but 2 of the 11 hospitals utilized nium bromide, and methylated spirit) bean Centre, Port of some apparatus to sterilize either the as compared to the other 7 institutions Spain, Trinidad. disinfectants/antiseptics or their con- that were involved in the question- The disk diffusion method (17) was tainers. These containers are used to naire component of the study. Of the 4 used to determine the antibiotic sensi- prepare or dilute the disinfectants/an- hospitals, 2 used pipeborne water as tivity of the isolates, on Mueller-Hin- tiseptics and are also used for distribu- well as boiled water for some prepara- ton agar. A total of 14 antimicrobial tion of disinfectants/antiseptics to the tions while the other 2 used distilled agents were used at the following con- pediatric/neonatal wards and the sur- water. centration: ceftriaxone (30 mg), strep- gical wards. Pharmacists were respon- Of the total of 180 tested samples tomycin (10 mg), ampicillin (10 mg), sible for preparing the disinfectants in 7 from the four hospitals, 11 samples cephalothin (30 mg), sulfamethoxa- of the 11 health institutions, and hospi- (6.1%) were contaminated by aerobic

Rev Panam Salud Publica/Pan Am J Public Health14(3), 2003 195 TABLE 1. Prevalence of aerobic bacteria in disinfectants/antiseptics by source in four hospitals in Trinidad, 2002

Samples contaminated No. of samples TAPC of contaminated samples Level/Type of sample tested No. % (CFU/mL)a

Pharmacy departments Pharmacy stock 36 0 0.0 NAb Dilutedcsample pre-pediatric/neonatal wardsd 36 3 8.3 3.4 2 103, <10,e<10e Diluted sample pre-surgery wardsf 36 3 8.3 3.2 2 104, <10,e<10e Pediatric/Neonatal wards Pre-use and in-use 36 3 8.3 2.95 2 103, 2.15 2 103, <10e Surgical wards Pre-use and in-use 36 2 5.5 3.6 2 104, 1.7 2 104 Total/Overall 180 11 6.1 NA aTotal aerobic plate count (TAPC) in colony-forming units per milliliter (CFU/mL). bNA = not applicable. cDiluted with distilled, pipeborne, or boiled water. dDiluted disinfectant/antiseptic samples destined for pediatric/neonatal wards. eSamples were negative for aerobic bacteria with the surface plating method but grew aerobic bacteria following enrichment (qualitative determination). fDiluted disinfectant samples destined for surgical wards.

bacteria (Table 1). Of the 11 contami- gluconate and cetrimonium bromide tals (19). Considerably higher preva- nated samples, 6 of them (54.5%) were samples, 4 of them were from the phar- lence levels have been reported for obtained in the pharmacy depart- macy departments, 3 were pre-use or other countries, including 34.4% in Ni- ments, and the other 5 (45.5%) were in-use samples from pediatric/neona- geria (7) and 43% in Japan (8). How- pre-use/in-use samples obtained from tal wards, and 2 were pre-use or in-use ever, when comparing the frequency the pediatric/neonatal wards or surgi- samples from surgical wards. of contamination, one should consider cal wards. As mentioned earlier, all the bacter- the types and concentrations of disin- Of the four health institutions stud- ial isolates found were identified as fectants since resistance varies among ied, two of them had contaminated Pseudomonas spp. The prevalence of re- microorganisms (2, 4). disinfectants/antiseptics. Of the three sistance of these 24 isolates to the 14 The total aerobic plate count that disinfectants/antiseptics studied, aero- antimicrobial agents is shown in Table we found, which ranged from under bic bacteria contaminated two samples; 3. Of the 24 isolates, all of them were 10 CFU/mL to 3.6 2 104 CFU/mL, is only the methylated spirit samples resistant to one or more antimicrobial similar to reports by others, where the were completely free of contamination. agents. The prevalence of resistance counts ranged from 102to 108CFU/ All the bacterial isolates found in the ranged from 25.0% (tetracycline) to mL (8, 9). The relatively high count of contaminated samples were identified 100% (ceftriaxone and streptomycin). aerobic bacteria in the samples is in- as Pseudomonas spp. For the contami- The differences in prevalence of resis- dicative of the probability of attaining nated samples, the sample with the tance to antimicrobial agents were sta- an infective dose and of establishing highest aerobic plate count, of 3.6 2 104 tistically significant (c 2test, P< 0.05). an infection (20, 21). The risk is further colony forming units per mL (CFU/ The patterns of resistance among the magnified when contaminated disin- mL), originated from a surgical ward. Pseudomonas spp. isolates were not sig- fectants are used in sensitive areas of Table 2 shows the prevalence of aer- nificantly different across the three the health facilities such as surgical obic bacteria found, by type of disin- levels (pharmacy departments, pedi- wards and pediatric/neonatal wards. fectant/antiseptic and by level. All 60 atric/neonatal wards, and surgical Clearly, any resulting nosocomial in- samples of methylated spirit studied wards). fections could have grave conse- were free of aerobic bacterial contami- quences (20, 22, 23). nants. However, 2 of the 60 samples of Of the three disinfectants/antisep- chlorhexidine gluconate (3.3%) were DISCUSSION tics investigated, chlorhexidine glu- positive for aerobic bacteria, and 9 of conate and cetrimonium bromide the 60 samples of chlorhexidine glu- The 6.1% prevalence of contami- (which was a 1% concentration) was conate and cetrimonium bromide were nated disinfectant samples from the the most contaminated, with 15% of positive. The contamination of the 2 four hospitals in Trinidad compares the samples positive for aerobic bacte- chlorhexidine gluconate samples oc- favorably with the 3% prevalence ria. Since similar practices took place curred in the pharmacy departments. found in 11 Danish hospitals (18) and in the preparation of the three disin- Of the 9 contaminated chlorhexidine the 7.9% reported for Malaysian hospi- fectants/antiseptics studied, it was

196 Gajadhar et al. • Microbial contamination of disinfectants and antiseptics in four major hospitals in Trinidad Rev Panam Salud Publica/Pan Am J Public Health14(3), 2003 197 expected that they would all have been to dilute the disinfectants/antiseptics have been reported to be the predomi- exposed to the same degree of contam- as well as failing to maintain adequate nant aerobic bacteria recovered from ination. However, the difference in the cleanliness of the disinfectant/antisep- contaminated disinfectants/antisep- frequency of contamination may be ex- tic containers are important sources of tics elsewhere (8, 19). It has also been plained in part by the low concentra- bacteria. Deionized water used in di- reported that different Pseudomonas tion (1%) of chlorhexidine gluconate luting chlorhexidine gluconate and spp. vary in their resistance to disin- and cetrimonium bromide, which ob- cetrimonium bromide was the source fectants (2, 5). However, bacteria other viously was unable to inhibit or kill the of P. maltophiliain a nosocomial epi- than Pseudomonashave been isolated contaminants. Given that this sub- demic in Australia (20). In Trinidad, from contaminated disinfectants and stance was the one that the 11 hospitals pipeborne water has been reported to have been implicated in nosocomial used most frequently to clean wounds, have a low residual level and epidemics (19, 29). the health risks for patients in Trinidad high coliform counts (25, 26), contrary Of therapeutic relevance is our find- are clear. Contaminated chlorhexidine to the recommended zero tolerance for ing that all 24 isolates of Pseudomonas gluconate and cetrimonium bromide coliforms in potable water (27). Never- spp. tested exhibited resistance to one was responsible for an outbreak of theless, the pipeborne water at these or more of the 14 antimicrobial agents Pseudomonas maltophilia amongst Aus- four hospitals did not appear to be the tested. In the healthcare delivery sys- tralian patients (20). main source of aerobic bacteria conta- tem in Trinidad and Tobago, in treat- Of the 60 chlorhexidine gluconate mination. This is because aerobic bac- ing Pseudomonas spp. infections, gen- samples in our study, only 2 of them teria were not detected in any of the tamicin is the empirical choice for (3.3%) were contaminated by aerobic disinfectant/antiseptic samples from intravenous treatment of inpatients bacteria. Nevertheless, the fact this the two of the four hospitals where who have good renal functions, disinfectant was applied on skin sur- disinfectant containers are subjected to whereas ceftazidime is prescribed for faces did not entirely eliminate the a high temperature while undergoing patients with renal impairment (30). It risk that it poses to the patients in two cycles each of wash and rinse, fol- is worrisome that 41.7% of the 24 iso- Trinidad. Other research (8) has re- lowed by autoclaving at 255 °F at 15 lb lates in our study were resistant to ported a considerably higher preva- per square inch of pressure (124 °C at gentamicin and that 100% of them lence, 60%, for contaminated chlorhex- 1.055 kg/cm2) for 7 to 10 min. Conta- were resistant to ceftriaxone. idine gluconate samples. minated containers can contaminate In Trinidad and Tobago one of the All the samples of methylated spirit disinfectants (21, 28). quinolones, ciprofloxacin, is the first that we studied from all the sources Studies elsewhere have shown that line of oral therapy for Pseudomonas were negative for aerobic bacteria. in-use disinfectants are more heavily infections. Of the 24 isolates in our This suggests that the concentration of contaminated than diluted samples at study, 58.3% of them were resistant to this antiseptic and its mode of activity the pharmacy level (7, 19, 20). That is ciprofloxacin, again emphasizing the completely eliminated all the contami- because the containers and the length possible therapeutic consequences of nating aerobic bacteria that we found of time that diluted samples are used infections that result from exposure to in the two other disinfectants that we in the wards affect contamination and contaminated disinfectants/antiseptics. studied. By its inherent nature, methy- the growth of bacteria. However, in In the high-risk areas of the hospi- lated spirit would not harbor microor- our study, we did not individually tals (the pediatric/neonatal wards and ganisms, and it is used in its original trace the disinfectant/antiseptic sam- the surgical wards) 5 out of the 72 concentration from the supplier. ples in a longitudinal fashion from the samples of pre-use or in-use disinfec- It was hardly a surprise that all stock pharmacy departments to the pedi- tants/antiseptics were contaminated. solutions of the three disinfectants/an- atric/neonatal wards or the surgical This 6.9% prevalence is obviously un- tiseptics tested in all the pharmacies wards of the hospitals; we took sam- acceptable because of the health risk were free of aerobic bacteria. This ples at the hospitals only at weekly in- posed to patients in these sensitive could be explained in part by the high tervals. This limitation may partially areas. Combined with the relatively concentration of the active ingredient explain our failure to detect a higher high counts of Pseudomonas spp. iso- in the disinfectants/antiseptics prior to prevalence or counts of aerobic bacte- lated, the high prevalence of resistance dilution, as well as the fact that they ria in the in-use disinfectant/antisep- to antimicrobial agents commonly had not yet been exposed to potential tics in the pediatric/neonatal wards used in their control is definitely a environmental contamination. and the surgical wards as compared source of concern. In order to mini- It has been established that unsani- with diluted disinfectant/antiseptic mize the risk posed to patients by con- tary practices during the preparation samples in the pharmacy departments taminated disinfectants/antiseptics in and distribution of disinfectants/anti- awaiting distribution in the hospitals. health facilities in Trinidad and To- septics contribute significantly to their It was noteworthy that Pseudomonas bago, we recommend that: contamination in a hospital environ- spp. accounted for all the aerobic bac- ment (10, 20, 22–24). For example, teria isolated from the disinfectant •Standardized guidelines for pre- using inappropriate sources of water samples studied. Pseudomonas spp. venting microbial contamination of

198 Gajadhar et al. • Microbial contamination of disinfectants and antiseptics in four major hospitals in Trinidad disinfectants/antiseptics be insti- •Health facilities should be encour- tine random microbiological sam- tuted and enforced in all hospital aged to use freshly prepared disin- pling of disinfectant/antiseptic sam- pharmacies. fectants/antiseptics and to institute ples before leaving the pharmacy •Functional autoclaves be provided measures to monitor (e.g., by expiry departments and while in use at by the Government to all public hos- dating) diluted disinfectants/anti- their destinations in the hospitals. pitals or health institutions that pre- septics to minimize prolonged use pare solutions such as disinfec- in the postdilution period; wherever Acknowledgements. We thank the tants/antiseptics for utilization in possible, single-use containers, with Pharmacy Programme of the Faculty of critical areas, e.g., pediatric/neonatal a maximum volume of 500 mL, Medical Sciences of the University of wards and surgical wards, in order should be introduced. the West Indies for funding the project. to sterilize all disinfectant/antiseptic •Systems should be instituted to de- We are grateful to Dr. I. Bekele and Ms. containers; in cases where the de- termine the microbial load of disin- Nadira Seepersadsingh for their sup- sired sterilization temperature is not fectants/antiseptics, particularly of port. Ms. Jan Rodriguez is acknowl- achievable, all containers must be in-use preparations, in order to min- edged for typing the manuscript. subjected to at least 100 °C prior to imize the potential health risk that filling with disinfectants/antiseptics. they pose. This can be done by rou-

REFERENCES

1.Newman KA, Tenney JH, Oken HA, Moody 12.Herruzo-Cabrera R, Uriarte MC, Rey-Calero J. 23.Fox JG, Beaucage CM, Folta CA, Thornton MR, Wharton R, Schimpff SC. Persistent isola- Antimicrobial effectiveness of 2% glutaralde- GW. Nosocomial transmission of Serratia tion of an unusual Pseudomonasspecies from a hyde versus other disinfectants for hospital marcescensin a veterinary hospital due to con- phenolic disinfectant system. Infect Control equipment, in an in vitro test based on germ- tamination by . J Clin 1984;5:219–222. carriers with a high microbial contamination. Microbiol 1981;14:157–160. 2.Whitmore E, McBee ML, Miner NA, Klasky S. Rev Stomatol Chir Maxillofac 1999;100:299– 24.Ojajarvi J. Effectiveness of and Susceptibility of Pseudomonasto disinfectant. 305. disinfection methods in removing transient Respir Care 1975;20:745–749. 13.Kajanahareutai S, Rahule S, Sirikulsatein P, bacteria after patient nursing. J Hyg (Lond) 3.Winnefield M, Richard MA, Drancourt M, Sangkasuwan S, Yospol P. Efficacy and conta- 1980;85:193–203. Grob JJ. Skin tolerance and effectiveness of mination of in-use disinfectants in Rajavithi 25.Welch P, David J, Clarke W, Trinidade A, Pen- two hand decontamination procedures in General Hospital. J Med Assoc Thai 1995;78 ner D, Bernstein S, et al. Microbial quality of everyday hospital use. Br J Dermatol 2000; Suppl 1:S36–S39. water in rural communities of Trinidad. Rev 143:546–550. 14.Ayliffe GA. Hospital disinfection and antibi- Panam Salud Publica 2000;8(3):172–180. 4.Bageant RA, Marsik FJ, Kellogg VA, Hyler otic policies. Chemioterapia 1987;6:228–233. 26.Agard L, Alexander C, Green S, Jackson M, DL, Gruschel DH. In-use testing of four glu- 15.Khor SY, Jegathesan M. In-use testing of dis- Patel S, Adesiyun AA. Microbial quality of taraldehyde disinfectants in the Cidematic infectants in Malaysian government hospitals. water supply to an urban community in washer. Respir Care 1981;26:1255–1261. Med J Malaysia 1977;32:85–89. Trinidad. J Food Prot 2002;65:1297–1303. 5.Russell AD. Bacterial resistance to disinfec- 16.Macfaddin JF. Biochemical tests for identifica- 27.World Health Organization. Guidelines for tants: present knowledge and future prob- tion of bacteria. New York: Williams and drinking-water quality. Health criteria and lems. J Hosp Infect 1999;43:S57–S68. Wilkins; 1980. other supporting information. Geneva: WHO; 6.Rutala WA. APIC guidelines for selection and 17.Baeur AW, Kirby WMM, Sherris JC, Turck M. 1993. use of disinfectants. 1994, 1995 and 1996 APIC Antibiotic susceptibility testing by a standard- 28.Anderson K, Keynes R. Infected corks clo- Guidelines Committee. Association for Pro- ized single disk method. Am J Clin Pathol sures and apparent survival of organisms in fessionals in Infection Control and Epidemiol- 1966;45:493–496. aseptic solutions. Br Med J 1958;2:27. ogy, Inc. Am J Infect Control 1996;24:313–342. 18.Christensen EA, Jepsen OB, Kristensen H, 29.Hsueh PR, Teng LJ, Yang PC, Pan HL, Ho SW, 7.Olayemi AB, Obayan Y. Contaminated disin- Steen G. In-use tests of disinfectants. Acta Luh KT. Nosocomial pseudoepidemic caused fectants in health clinics in Ilorin, Nigeria. In- Pathol Microbiol Immunol Scand [B] 1982;90: by Bacillus cereustraced to contaminated ethyl fect Control Hosp Epidemiol 1994;15:581–582. 95–100. alcohol from a liquor factory. J Clin Microbiol 8.Oie S, Kamiya A. Microbial contamination of 19.Keah KC, Jagathesan M, Tan SC, Chan SH, 1999;37(7):2280–2284. antiseptics and disinfectants. Am J Infect Con- Chee OM, Cheong YM. Bacterial contamina- 30.Herfindal ET, Gourley DR, eds. Textbook of trol 1996;24:389–395. tion of hospital disinfectants. Med J Malaysia therapeutics: drug and disease management. 9.Zembrzuska-Sadkowska E. Microbial conta- 1995;50:289–290. Seventh ed. Philadelphia: Lippincott Williams mination of pharmacies and hospital wards 20.Wishart MM, Riley TV. Infection with and Wilkins; 2000. environment and its influence on the purity of Pseudomonas maltophiliahospital outbreak due prescription preparations during their pro- to contaminated disinfectant. Med J Austr duction process as well as their application. 1976;2:710–712. Acta Pol Pharm 1995;52:67–75. 21.Jones BL, Gorman LJ, Simpson J, Curran ET, 10.Keah KC, Jegathesan M, Tan SC, Chan SH, McNamee S, Lucas C. An outbreak of Serratia Chen OM, Cheong YM, et al. Disinfection: marcescensin two neonatal intensive care gaps between recommended and actual prac- units. J Hosp Infect 2000;46:314–319. tice. Southeast Asian J Trop Med Public 22.Zembrzuska-Sadkowska E. The danger of in- Health 1995;26:795–798. fections of the hospitalized patients with the 11.Rutala WA, Weber DJ. Surface disinfection: microorganisms present in the preparations should we do it? J Hosp Infect 2001;48: and in the hospital environment. Acta Pol Manuscript received 19 September 2002. Revised version S64–S68. Pharm 1995;52:173–178. accepted for publication on 3 March 2003.

Rev Panam Salud Publica/Pan Am J Public Health14(3), 2003 199 RESUMEN Objetivo. Evaluar la contaminación microbiana de los desinfectantes y antisépticos usados en cuatro grandes hospitales de la isla caribeña de Trinidad. Métodos. En este estudio de tipo transversal se tomaron muestras de desinfectantes La contaminación microbiana y antisépticos obtenidos de los departamentos de farmacia, de las salas de pediatría y de desinfectantes neonatología y de las salas de cirugía de cuatro hospitales. Las muestras fueron culti- y antisépticos en cuatro vadas en medio de agar nutritivo para aerobias mediante la técnica de siem- bra superficial. La sensibilidad de las bacterias aisladas a los antibióticos se determinó grandes hospitales por el método de difusión en placa usando 14 antimicrobianos distintos. En total es- de Trinidad tudiamos 180 muestras de desinfectantes o antisépticos: 60 de gluconato de clorhexi- dina (Hibitane), 60 de gluconato de clorhexidina más bromuro de cetrimonio (Savlon) y 60 de alcohol metílico. Resultados. De las 180 muestras estudiadas, 11 (6,1%) estaban contaminadas con bacterias aerobias, invariablemente Pseudomonasspp. De las 11 muestras contami- nadas, 6 (54,5%) provenían de farmacias, mientras que 5 (45,5%) correspondían a muestras de producto diluido obtenidas antes de usarse o mientras se usaba el pro- ducto en las salas pediátricas, neonatales y quirúrgicas del hospital. Nueve de los 11 (81,8%) desinfectantes o antisépticos contaminados fueron una mezcla de gluconato de clorhexidina más bromuro de cetrimonio, y los 2 restantes (18,2%) fueron solo glu- conato de clorhexidina. Únicamente dos de los cuatro hospitales tuvieron muestras de desinfectantes o antisépticos contaminadas. Cada uno de los 24 aislamientos de Pseudomonasspp. mostró resistencia a uno o más de los 14 antimicrobianos de la prueba, y las prevalencias observadas de resistencia a ciprofloxacina, norfloxacina, to- bramicina y gentamicina fueron de 58,3%, 50,0%, 45,8% y 41,7%, respectivamente. Conclusiones. Según estos resultados, la contaminación de desinfectantes o antisép- ticos constituye un peligro para la salud de los pacientes, sobre todo en las salas pediátricas y quirúrgicas. La alta prevalencia de resistencia a antimicrobianos que se observa en las Pseudomonas spp. aisladas en este estudio es especialmente alarmante.

200 Gajadhar et al. • Microbial contamination of disinfectants and antiseptics in four major hospitals in Trinidad