Microbial Prevalence in Domestic Humidifiers - Applied and ...
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Apr. 1980, p. 840-844 Vol. 39, No. 4 0099-2240/80/04-0840/05$02.00/0 Microbial Prevalence in Domestic Humidifiers HARRIET A. BURGE,* WILLIAM R. SOLOMON, AND JEAN R. BOISE Allergy Division, Department of Internal Medicine, University of Michigan, Ann Arbor, Michigan 48109 The prevalence of viable thermophilic bacteria and actinomycetes and meso- philic fungi was examined in 145 samples from 110 domestic humidifiers. A total of 72 and 43% of furnace and console humidifier samples, respectively, contained viable thermophilic bacteria, whereas 60 and 72% of these samples produced Downloaded from http://aem.asm.org/ on January 26, 2021 by guest mesophilic fungal growth. Thermophilic actinomycetes were recovered from seven humidifier samples. Efforts to detect thermophilic actinomycete antigens in 15 humidifier fluid samples were not successful. Antifoulants added to humi- difier fluid reservoirs had no apparent effect on microbial frequency. Airbome microbial recoveries did not reflect patterns of humidifier contamination with respect to either kinds or numbers of microorganisms in 20 homes in which volumetric air samples were obtained during humidifier operation. Like many other microorganisms, thermo- (Becton, Dickinson & Co., Rutherford, N.J.), a Zip- philic actinomycetes are known to colonize home lock plastic bag, and a form on which they provided humidification systems. Thermoactinomyces the following information: name, address, telephone species were isolated from 17 of 44 humidifier number, date sample taken, age of home, years occu- pied by subject, heating type and fuel, dwelling type samples (39%) as well as from a number of other and construction; indoor conditions (i.e., very humid, household substrates by Kurup et al. (11) and humid, comfortable, rather dry, parched); and make from 12 of 20 humidifier water samples by Fink and model number of humidifier, type, capacity, how et al. (9). Banazak et al. (1) also reported acti- long in use in current season, date(s) and type(s) of nomycete contamination in an unspecified num- most recent cleaning, antifoulants used if any, and ber of humidifiers as well as in 50 to 75% of home location (consoles). Patients were asked to fill the air samples. Additional reports have dealt with syringe with humidifier fluid, replace the plastic screw specific humidifiers implicated in cases of hyper- cap on the syringe hub, place the syringe in the plastic sensitivity pneumonitis from which these micro- bag, and refrigerate the sample, bringing it to their next (weekly or biweekly) clinic visit. organisms were isolated (4, 8, 12, 19). These This procedure produced 63 patient-collected sam- documented recoveries have stimulated a justi- ples in 1975-1976 and 46 in 1976-1977. Samples col- fiable concern over the possible risks of home lected by laboratory personnel during home visits (10 humidifying systems especially for people with in 1977, 28 in 1977-1978) included a 50-ml sample of respiratory disease. undisturbed humidifier fluid, a 20-ml sample of hum- The studies to be presented here were de- idifier fluid taken after vigorous scraping of internal signed to provide a quantitative survey of micro- reservoir surfaces, and a 50-ml sample of tap water bial prevalence in domestic humidifying units, from the home. In addition, fluid temperature and pH to assess determinants of such contamination, measurements (pHydrion Papers, Microessential Lab- and to examine the potential for aerosolization oratories, Brooklyn, N.Y.) were made in situ, and the same information was elicited as for the patient-col- of viable microbial particles from these potential lected samples. Home visits also permitted air sam- sources. pling to be done during humidifier operation. Single MATERLALS AND METHODS plates of malt extract agar (32 g of malt extract, 16 g of agar, 1 g of yeast extract [all from Difco Laborato- Humidifiers for study were identified by approach- ries, Detroit, Mich.], and 1,000 ml of distilled water) ing patients who received immunotherapy injections were simultaneously exposed beneath the sixth stage for respiratory allergy at the University of Michigan of each of two upright Andersen samplers (Andersen allergy clinics. These subjects, all of whom lived within Samplers, Inc., Atlanta, Ga.) as previously described 50 miles of Ann Arbor, were familiarized with the (16). Air was drawn into the samplers at a rate of 0.03 objectives of the study and asked for their cooperation. m3/min for 1 min. Subsequently, single plates of Tryp- Volunteers initially collected a sample of humidifier ticase soy agar (BBL, Cockeysville, Md.) were simi- reservoir fluid which was brought to the clinic (1975- larly exposed. Air temperature and relative humidity 1976, 1976-1977). In the second and third seasons of measurements were made with a sling psychrometer the study (1977 and 1977-1978), a similar patient group during these sampling procedures. agreed to an in-home study conducted by allergy re- A 0.1-ml portion of each humidifier sample was search staff. Patients volunteering to collect their own cultured on Trypticase soy agar, Trypticase soy agar samples were given a sterile 50-ml disposable syringe with 30 Lg of novobiocin per ml (Sigma Chemical Co., 840
VOL. 39, 1980 MICROBIAL PREVALENCE IN DOMESTIC HUMIDIFIERS 841 St. Louis, Mo.) and malt extract agar with 0.8 ug of never effected automatically. A total of 36 of the chlortetracycline (Sigma). Trypticase soy agar cul- furnace humidifiers (FH) were of the rotating- tures were incubated at 50°C in a humidified incuba- drum type, having a circular belt of plastic mesh tor; malt extract agar was incubated at room temper- which is passed continually through a fluid col- ature. Samples that produced confluent growth on either medium later were diluted 1:1, 1:2, and 1:4 and lection. The belt is applied to a drum which recultured. Humidifier solids recovered after scraping revolves in the air stream to be moistened. Com- were scattered on the surface of Trypticase soy agar monly, a sump was provided to prevent overfill- with novobiocin and incubated at 50°C. ing of the unit. Thirteen FH were of the tric- Colony counts were expressed as colony-forming kling-column type in which fluid is pumped to units (CFU) per milliliter of fluid or CFU per cubic the top of a rectangular screen of coarse metal meter of air. Bacteria were Gram stained and classified fibers, through which it then percolates. Air only by shape and Gram stain reaction. Thermophilic passing through the mesh is moistened in an Downloaded from http://aem.asm.org/ on January 26, 2021 by guest actinomycetes were identified by the method of Kurup and Fink (10). Mesophilic fungi were identified with operation that effectively obviates a pool or res- the aid of keys and descriptions in Barnett and Hunter ervoir of fluid. The types for the remaining eight (2), Barron (3), and Ellis (7). FH were not specified. For double immunodiffusion studies of humidifier Temperature and pH data and water quality fluid (13) glass microslides coated in Noble agar measurements for the humidifier fluids (Table (Difco) were employed. Center wells were filled with 2) were obtained only on the 38 occasions where 0.01-ml aliquots of humidifier fluid; antisera were home air samples also were obtained. Microbial placed in outer wells in dilutions from 1:1 to 1:128 recoveries from all humidifier samples are sum- (Greer Laboratories, Lenoir, N.C.). A 1:1 mixture of marized in Table 3. FH were more frequently humidifier fluid and an appropriate antigen (Greer) free of microorganisms detectable in our system was used in central wells on control slides. Antisera used were Thermoactinomyces vulgaris, Thermoac- than CH. Mesophilic fungi were recovered from tinomyces candidus, Saccharomonospora viridis, and CH with slightly greater frequency than from Micropolyspora faeni. FH, whereas the reverse was true for thermo- Percent dry weight was measured by oven drying in philic bacteria. Thermophilic actinomycetes preweighed containers of 2 ml of humidifier fluid. were recovered from only four FH and three Hardness, calcium, nitrate, phosphate, and silica were CH. Of the fluid samples from the 36 rotating- measured with water quality testing kits (Ecologic, drum FH, 17% were sterile compared with 31% Bohemia, N.Y.). of 13 trickling-column FH units. A total of 70 and 83% of the rotating-drum units produced RESULTS mesophilic and thermophilic cultures, respec- Categories of humidifier fluid and concurrent tively, whereas 46% of the trickling-column FH home air samples taken are summarized in Table were similarly contaminated. 1 by humidifier type. All the console humidifiers Antifoulants apparently had no effect on (CH) were of a similar design in which a fan either the frequency or level of microbial con- blows air past a felted fiber belt moistened by tamination in either humidifier type (Table 3). repeated passage through a shallow reservoir of The most frequently used antifoulant was house- fluid. Water is added manually from above and, hold bleach (four CH, one FH). Sixteen other in general, drainage of the unit is difficult and treatments were used, no one in more than three humidifiers. TABLE 1. Summary of humidifier and associated Fifteen FH and nine CH were sampled more home air samples than once (Table 1). Most of these repeat sam- ples followed a period of nonuse during which No. of the humidifiers were cleaned. Thermophiles samples were recovered two or more times in nine FH Furnace and in three CH and were consistently absent in Various types sampled 58 four FH and four CH. Mesophiles were consist- Sampled twice ......... ......... 8 ently present in seven FH. Of these, the taxa Sampled three times .. 6 recovered were consistent in three. Mesophiles Sampled four times ...... .............. 1 were consistently present in three CH, two of Total samples ............ 80 Associated home air samples .30 which were consistent with respect to taxa re- covered. Console Thernophilic microorganisms recovered from Various types sampled ... 53 humidifier fluids were primarily Bacillus spe- Sampled twice ......... 6 cies. In addition, Saccharomonospora viridis Sampled three times . 3 and T. vulgaris were recovered from two and Total samples .65 four humidifiers, respectively. To test the pos- Associated home air samples. 8 sibility of inhibitors in the humidifier fluids or
842 BURGE, SOLOMON, AND BOISE APPL. ENVIRON. MICROBIOL. TABLE 2. Water quality parameters measured on 30 furnace and 8 console humidifier samples Parameter SampleNirt from: Temp Hardness (mg/ Calcium Nitrate Phosphorus Silica Dry wt (mg/li- (0C) liter) (mg/li ter) (mg/liter) (mg/liter) ter) Console humidifer 13-17a 6-10 75-1,035 50-475 0-440 0.1-1.0 10-25 700-19,000 14.7b 7.5 537 244 114 0.4 15 8,900 15C 6.5 440 175 53 0.15 12.5 4,800 Fumace Downloaded from http://aem.asm.org/ on January 26, 2021 by guest humidifier 14-29 4-10 150-12,350 0-6,000 0-440 0.1-10.0 5-60 400-206,900 21 7 4,316 1,206 118 1.1 29 34,000 21 7.2 2,200 570 88 0.3 20 12,000 Tap water 100-160 50-95 0-2.6 0.1-1.0 1.5-10 131 63 1.4 0.17 2.6 145 55 1.8 0.1 5.5 a Range. b Mean. 'Median. TABLE 3. Frequency of microbial contamination in TABLE 4. Mesophilic taxa recovered from humidifier fluid samples humidifier fluids Console Furnace Humidifier fluid na "Ster- % Me- Ther- % Acti- sample ile" b sophilic mo- cetes Mesophilic taxa % Posi- Range % Posi- Range ~~philic ~~ tive (colonies/ tive (colonies/ Furnace total 57 23 60 72 7 With antifoulant 10 20 60 70 Alternaria 3.5 10 Without 47 23 60 72 Aspergillus 3.8 10-30 3.5 10 antifoulant Aureobasidium 1.9 10 Console total 53 11 72 43 6 Candida 1.9 10 With antifoulant 21 9 76 52 Cladosporium 5.7 10-80 10.5 10-970a Without 32 13 69 38 Cylindrocarpon 1.7 10 antifoulant Epicoccum 1.9 10 Fusarium 5.7 10-120 15.8a 10-590a 'Number of humidifier fluid samples. Gliocladiuma 3.8 10-60 3.5 20-260a b With no microorganisms recovered in our culture system. 10-80 7.0 10-20 Penicillium 9.4 Phialophoraa 15.1a 10-2,180a 15.8a 30-1,260a Rhodotorula 11.3a 10-100a 14.00 10-1,230a in (local) Ann Arbor tap waters or both, T. Scolecobasidium' 3.5 20-3,000a vulgaris and T. candidus were inoculated into Sepedonium' 1.7 980a Sordaria 1.9 20 five apparently sterile humidifier samples and Sphaeropsidales 1.9 40 two samples of tap water. Both organisms were Sporobolomyces 1.7 20 readily recovered from all inoculated samples. Verticillium 3.5 10-20 Fifteen humidifier fluid samples were tested by Yeast 28.30 10-2,5000 7.0 10-870a double immunodiffusion with antisera to T. can- Nonsporulating 1.9 10 5.3 10-1,500a didus, T. vulgaris, S. viridis, and M. faeni. No a Refers to unusual taxa, frequency, or level. humidifier fluid produced precipitin lines against any of the antisera. Mesophilic recoveries from these homes, similar taxa were encountered in humidifier fluids are summarized in Table 4. the air and the humidifier samples. In none of Spearman rank order correlations between these these cases could air contamination by humidi- mesophilic recoveries and other humidifier fluid fier organisms be clearly demonstrated since all parameters (temperature, pH, percent dry taxa recovered in both fluid samples and in air weight, hardness) were not significant. are commonly present in both indoor and out- In 17 homes in which air samples were made door air in our area. None of the unusual humi- concurrently with humidifier samples, the hum- difier organisms (e.g., Phialophora) were en- idifiers were contaminated with unusual num- countered in room air. Thermophilic actinomy- bers or types of organisms (Table 5). In seven of cetes (Thermomonospora fusca, T. candidus,
VOL. 39, 1980 MICROBIAL PREVALENCE IN DOMESTIC HUMIDIFIERS 843 TABLs 5. Mesophilic recoveries from 17 humidifier fluid and associated air samples Humidifier fluid Air Pink yeast, Phialophora Penicillium, pink yeast Pink yeast Penicillium, Aspergillus, Alternaria, Cladosporium, arthrospores Bacteria, pink yeast Penicillium, Aspergillus, Cladosporium, Alternaria Bacteria, Cladosporium Aspergillus, Penicillium, pink yeast Fusarium Penicillium, Sporobolomyces, Aspergillus, Cladosporium Scolecobasidium, Fusarium, Penicillium, Aspergillus, pink yeast Alternaria Fusarium, nonsporulating, Pink yeast, Penicillium, Cladosporium, Aspergillus arthrospores Downloaded from http://aem.asm.org/ on January 26, 2021 by guest Nonsporulating, Cladosporium, Aspergillus, Wallemia, Cladosporium, pink yeast, Penicillium, Fusarium, Aspergillus, bacteria Myriococcum, Olpitrichum, Epicoccum Phialophora, Fusarium, Penicillium Aspergillus, Penicillium, nonsporulating, yeast Fusarium, yeast Aspergillus, Penicillium, pink yeast, Cladosporium Pink yeast, Verticillium, Aspergillus, pink yeast, Penicillium, Cladosporium, nonsporulating Tritarachium Phialophora Penicillium, Cladosporium, pink yeast, Alternaria, Aspergillus, Epicoccum Cladosporium, Scolecobasidium Penicillium, Aspergillus, pink yeast, nonsporulating Bacteria, Penicillium Aspergillus, Penicillium, pink yeast Yeast, pink yeast Nonsporulating, yeast, Cladosporium, Penicillium, Alternaria, arthrospores Fusarium, Penicillium Penicillium, Cladosporium, yeast, Aureobasidium, Candida, sphaeropsids, Alternaria Cylindrocarpon Cladosporium, Alternaria, yeast, Aspergillus, Aureobasidium, Epicoccum, Penicillium Streptomyces sp., S. viridis) which were rare (or idifier samples they analyzed. The health effects absent) in humidifier fluids were recovered from of these saprophytes remain unknown, as does the air of all homes studied in levels from 50 to the potential for exposure to pathogenic bacteria 5,000/m3 of air sampled. from domestic humidifying systems. Temperature conditions in CH and most FH DISCUSSION were well within the optimum range for growth These studies, which admittedly are based on of mesophilic fungi, and these organisms were a circumscribed geographic area, suggest that recovered from 72 and 60% of such units. Many humidifiers, in fact, do not commonly contribute of the taxa recovered are common in either thermophilic actinomycetes to home air. In spite indoor or outdoor air in our area (e.g., species of of considerable effort to detect contaminating Penicillium and Cladosporium). However, Phi- thermophilic actinomycetes, we were able to alophora, which we have not isolated from air recover actinomycetes from only seven humidi- (18), was recovered from 15% of all humidifying fiers or 6% of those studied. On the other hand, units in levels ranging above 2,000 CFU/ml of these organisms were recovered from the air of fluid. Other fungi recovered from humidifier 100% of homes studied. Disease-provoking anti- fluid which are rarely isolated from air in our gens could, of course, be present in fluids unas- area are Sepedonium, Scolecobasidium, and sociated with currently viable microorganisms Gliocladium. The allergenic potential of these or where organisms are growing on inaccessible organisms is unknown, although Phialophora surfaces, as reported by Miller (12). However, species are known to cause invasive human dis- such antigens were not detected in the 15 fluids ease by inoculation into peripheral tissues. we tested. The risks of air contamination during opera- The maximum reservoir temperature meas- tion of contaminated humidifiers are not clear. ured in any humidifier was 290C, well below the Mist generators have been shown clearly to pro- 45 to 600C range optimal for thermophilic acti- duce viable particle aerosols (6, 17). Bacterial nomycetes. Yet, thermophilic bacteria with a emissions from evaporation-type humidifiers similar optimum temperature range were iso- (including all units reported here) also may oc- lated from 72% of our FH units and 43% of our cur when viable cells apparently are shaken from CH units. Seabury et al. (15) also reported large moist surfaces of fibrous material; however, re- numbers of thermophilic bacteria from all hum- sulting airbome levels apparently fall rapidly
844 BURGE, SOLOMON, AND BOISE APPL. ENVIRON. MICROBIOL. after turn-off (6). Schultz et al. (14) reported or both, the continued use of such systems seems that clearly contaminated evaporative humidi- defensible. fiers did not routinely produce bacterial aerosols, and our failure to demonstrate manifest airborne ACKNOWLEDGMENTS contamination argues against extensive intro- This work was supported by Public Health Service grant duction of viable microbes into room air during AI 10181 from the National Institute of Allergy and Infectious operation of the units studied. A potential for Diseases. acute exposure during cleaning does exist. LITERATURE CITED Cartwright and Aargrave (5) discussed humidi- fier scale as a nidus for humidifier microbial 1. Banazak, E. F., J. Barboriak, J. Fink, D. Schlueter, contamination. This material often is scraped, A. Sosman, W. Thiede, and G. Unger. 1974. Epide- miologic studies relating thermophilic fungi and hyper- scrubbed, and otherwise violently disturbed dur- Downloaded from http://aem.asm.org/ on January 26, 2021 by guest sensitivity lung syndromes. Am. Rev. Respir. Dis. 110: ing cleaning; hosing also may spread contami- 585-591. nated droplets into the air. 2. Barnett, H. K., and B. B. Hunter. 1972. Illustrated Antifoulants that had been added to the fluid genera of imperfect fungi, third edition. Burgess Pub- lishing Co., Minneapolis. reservoirs in 17% of FH and 40% of CH studied 3. Barron, G. L. 1968. The genera of hyphomycetes from did not produce clearly evident effects on the soil. The Williams & Wilkins Co., Baltimore. microbial flora. Covelli et al. (6) studied the 4. Burke, G. W., C. B. Carrington, R. Strauss, J. Fink, impact of a sodium hypochlorite antifoulant and and E. Gaensler. 1977. Allergic alveolitis caused by home humidifiers. J. Am. Med. Assoc. 238:2705-2708. reported that changes in Pseudomonas popula- 5. Cartwright, Y., and P. R. Aargrave. 1970. Pseudomo- tions lasted, at most, 24 h. By asking our vol- nas in ventilators. Lancet i:40. unteers not to disturb their humidifiers within 6. Covelli, H. D., J. Kleeman, J. Martin, V. Landau, and 24 h of our visit, we virtually insured that eva- R. Hughes. 1973. Bacterial emission from both vapor and aerosol humidifiers. Am. Rev. Respir. Dis. 108: nescent effects of antifoulants would not be no- 698-701. ticed. However, a more sustained influence of 7. Ellis, M. B. 1971. Dematiaceous hyphomycetes. Common. additives, in general, on microbial populations Mycol. Inst. Kew, Surrey. also was nowhere evident. These findings, com- 8. Fink, J. N., E. Banazak, W. Thiede, and J. Barboriak. 1971. Interstitial pneumonitis due to hypersensitivity to bined with possible adverse health effects of an organism contaminating a heating system. Ann. In- aerosolizing antifoulants, do not suggest that the tern. Med. 74:80-83. addition of currently available agents to humi- 9. Fink, J., A. Resnick, and J. Salvaggio. 1971. Presence difier reservoirs is desirable. of thermophilic actinomycetes in residential heating systems. Appl. Microbiol. 22:730-731. Long-term effects of physical cleaning on 10. Kurup, V. P., and J. N. Fink. 1975. A scheme for the humidifier microbial contamination are not well identification of thermophilic actinomycetes associated documented. Covelli et al. (6) reported that thor- with hypersensitivity pneumonitis. J. Clin. Microbiol. 2:55-61. ough cleansing and disinfection (glutaraldehyde) 11. Kurup, V. P., J. N. Fink, and D. M. Bauman. 1976. followed by dry storage resulted in only 24 h of Thermophilic actinomycetes from the environment. sterility after resumption of operation. However, Mycologia 68:662-666. only one-third of the CH examined more than 12. Miller, M. M. 1976. Chronic hypersensitivity lung disease once in our study were consistently contami- with recurrent episodes of hypersensitivity pneumonitis due to a contaminated central humidifier. Clin. Allergy nated with either thermophilic or mesophilic 6:451-462. microorganisms. All of these units were thor- 13. Ouchterlony, 0. 1968. Handbook of immunodiffusion oughly cleaned and stored dry between the two and immunoelectrophoresis. Ann Arbor Science Pub- samples and generally were cleaned several 14. Schultz, lishers, Inc., Ann Arbor. times during the period of use. On the other T., E. B. Edmondson, A. K. Pierce, and J. P. Sanford. 1967. Studies of a new humidifying device as hand, 60% of the FH sampled twice were con- a potential source of bacterial aerosols. Am. Rev. Res- sistently contaminated with thermophiles and pir. Dis. 96:517-519. 46% were contaminated with mesophiles. These 15. Seabury, J., B. Becker, and J. Salvaggio. 1976. Home units were not always cleaned between the two humidifier thermophilic actinomycete isolates. J. Al- lergy Clin. Immunol. 57:174-176. samples and were rarely cleaned during the pe- 16. Solomon, W. R., and J. A. Gilliam. 1970. A simplified riod of use. At least one FH and several CH, application of the Andersen sampler to the study of which were cleaned frequently and kept com- airborne fungus particles. J. Allergy 45:1-13. pletely free of scale, were consistently sterile by 17. Solomon, W. R. 1974. Fungus aerosols arising from cold- mist vaporizers. J. Allergy Clin. Immunol. 54:222-228. our cultural criteria. 18. Solomon, W. R. 1975. Assessing fungus prevalence in Given the low frequency of viable thermo- domestic interiors. J. Allergy Clin. Immunol. 56:235- philic actinomycete recoveries from the humi- 242. difying systems studied and the fact that the 19. Tourville, D. R., W. I. Weiss, P. T. Wertlake, and G. M. Leudemann. 1972. Hypersensitivity pneumonitis presence of such organisms does not necessarily due to contamination of home humidifier. J. Allergy imply respiratory exposure or clinical relevance Clin. Immunol. 49:245-251.
You can also read