374

NTRODUCTION

Bacterial contamination of treated water is a matter of

serious concern in hemodialysis (HD) centers and im-

provement of microbiological quality is a permanent

challenge for professionals at these units.

Patients on regular HD treatment are exposed to hun-

dreds of liters of dialysis water every week through the

extracorporeal circulation (1). Pyrogenic reactions

and bacteremia are potential consequences of poor

microbiological quality of treated water for the prepa-

ration of dialysate and reprocess in HD (2)

Several factors favour bacterial growth in the water

treatment systems for HD, including sand and carbon

filters and softeners, or parts with low circulation such

as tanks and taps (3). Formation of biofilms that facili-

tate bacterial persistence at different points of the sy-

stem and protect bacteria from disinfection also in-

creases the risk of contamination and high endotoxin

levels in water (4).

Despite continuous efforts at these centers to improve

the microbiological quality of treated water for HD,

water microbiological quality parameters must be con-

stantly assessed in order to avoid undesirable conse-

quences and implement improvement programs.

We report the results of a three-year study of the bac-

teriological quality of water samples in 32 out of a total

of 38 HD centers throughout Uruguay. All the sam-

ples were processed at our private independent labo-

ratory, as part of the centers' quality assurance pro-

gramme. The results had to be reported annually to a

special commitee, acting on behalf of the health

authorities. For this study, samples were obtained at

different points of the purification process and exami-

ned for heterotrophic bacteria, total coliforms, and

Pseudomonas aeruginosa

J N

EPHROL 2002; 15: 374-379

www.sin-italia.org/jnonline/vol15n4/

Microbiological quality of hemodialysis

water in a three-year multicenter study

in Uruguay

Pablo Zunino, Lucía Beltrán, Laura Zunino, Hernán Méndez, Verónica Percovich, Rossana Rocca, Beatriz

Antonelli

Beltrán - Zunino Laboratory (B-Z SRL), Montevideo - Uruguay

ABSTRACT:

Background:

The microbiological quality of treated water is a very important issue in hemodialysis (HD)

centers. Water treatment may have a dramatic effect on microbial contamination because of bacterial colonisation of

the different parts of the system such as storage tanks, softeners or deionisers. Therefore, HD centers must have

stringent quality programmes including regular water monitoring for microbiological analysis. We report the results

of a three-year study (July 97 to June 2000) including bacteriological quality analysis of 3129 water samples from 32

out of 38 HD centers throughout Uruguay.

Methods:

Bacteriological analysis of water samples was based on heterotrophic count, total coliform count and

Pseudomonas aeruginosa

presence/absence, according to the procedures proposed by the Association for the Ad-

vancement of Medical Instrumentation (AAMI).

Results:

Heterotrophic counts of 83% of the samples were under 200 colony forming units (CFU)/mL (AAMI limit of

compliance) and the water samples after the final bactericidal treatment showed 99% compliance. The points show-

ing the worst results were softeners and deionisers (60% acceptance).

Conclusions:

In comparison with a similar study in Uruguay, from January 96 to June 97 at the same microbiological

laboratory, there has been a marked improvement in the microbiological quality of water for hemodialysis.

Key words:

Water, Hemodialysis, Microbiological quality, Uruguay

Page 2

Zunino et al

375

UBJECTS AND METHODS

Water sampling.

Water was sampled monthly at most of

the centers, as recommended by the Association for

the Advancement of Medical Instrumentation (AA-

MI) (5) over a three-year period in 32 centres from va-

rious regions in Uruguay (July 1997 to June 2000).

Sampling was done either by unit workers or by labo-

ratory workers, depending on the dialysis center.

Water was sampled aseptically according to the Ameri-

can Public Health Association (APHA) methods (6).

About 120 mL of water were collected in commercial

sterile polypropylene bottles (Bioster

) after 1 to 2 mi-

nutes of free flow through the faucets.

Water sampling sites were as follows: sand filter, activa-

ted carbon filter, softener or deioniser, reverse osmo-

sis, treated water storage tank, HD room, membrane-

washing room (where dialysis membranes are washed

and treated for reuse), recirculation loop, and post re-

verse osmosis bactericidal treatment area (UV light or

ozone). In all HD centers, water samples obtained

from the HD room faucet or from the nearest availa-

ble sampling point were considered as treated.

In order to compare the performance of individual

HD centers, only those with at least 12 treated water

samples during the whole period and at least four in

the last year were considered.

A total of 3129 water samples were analysed during the

three years of the study.

Bacteriological analysis.

All bacteriological tests were do-

ne according to the Standard Methods for the Exami-

nation of Water and Wastewater (6). All bacteriologi-

cal media were from Difco, Detroit, Mi. (USA).

Heterotrophic bacteria counts were done in triplicate

on Plate Count Agar by the spread plate technique

and plates were incubated for 72 h at 35° C. Total co-

liforms were determined by membrane filtration: 100

mL of water was filtered using 0.45 µm pore size mem-

branes (Pall Gelman Laboratory). Membranes were

placed on m-Endo plates and incubated for 24 h at 35º

C. Coliforms were detected by production of typical

red colonies with a metallic surface sheen or atypical

dark red colonies without sheen. All were confirmed

for gas production in lauryl tryptose broth tubes, in-

cubated at 35º C for 48 h.

Pseudomonas aeurginosa

was also assessed: 10 mL of wa-

ter were added to asparagine broth and incubated at

35ºC for 48 h. Tubes showing turbidity and produc-

tion of fluorescent pigment, assessed with a long-wave-

length UV light lamp, were subcultured in Cetrimide

Milk Agar for determination of

P. aeruginosa

(6).

Bacteriological acceptance limits.

Standards for microbio-

logical water quality recommended by the AAMI were

used to determine compliance of water samples, ac-

cording to national health authority criteria (5). The-

se guidelines establish an acceptable limit of 200 co-

lony-forming units (CFU)/mL of heterotrophic bacte-

ria for treated water used for HD and 2000 CFU/mL

for dialysate.

ESULTS

Heterotrophic bacteria counts from all sampling sites.

Ac-

ceptable heterotrophic bacteria counts (<200

CFU/mL) of all samples varied from 60%, in the sof-

tener or deioniser samples, to 99% after bactericidal

treatment (UV light or ozone) (Tab. I).

Heterotrophic bacteria counts from samples of treated water.

When heterotrophic bacteria counts in treated water

were considered during the three-year period, 83%

of the samples were below 200 CFU/mL, 9%

between 200 and 1000 CFU/mL and 8% were over

1000 CFU/mL. When results were analysed for each

year of the study, an improvement in bacteriological

quality was observed since the percentage of samples

with counts below 200 CFU/mL was 73 in the first

year, 87 in the second and 89 in the third.

Heterotrophic bacteria counts of samples from HD room fau-

cets.

Heterotrophic bacteria counts of samples from

HD room faucets improved with time. The percenta-

ge of acceptable samples in the first year was 58, ri-

sing to 78 in the second, and 87 in the third and final

year.

Heterotrophic bacteria counts related to sampling month.

Geometric means of heterotrophic bacteria counts

in samples from HD room faucets were analysed in

relation to the collection month in order to assess

the influence of ambient temperature. During the

winter months (June to September) counts were no-

tably lower than over the rest of the year, when am-

bient temperature is higher (Fig. 1). The correlation

between the geometric mean of heterotrophic bacte-

ria counts and the mean temperature was 0.7, confir-

ming that temperature has an important impact on

microbiological water quality.

Fig. 1 - Geometric mean of heterotrophic counts (CFU/mL) at

hemodialysis room and ambient temperature (ºC) related to

sampling month, expressed as means for the three years.

CFU/mL (geometric mean)

Temperatur

e (°C)

Month

Page 3

Hemodialysis water in Uruguay

376

Heterotrophic bacteria counts of treated water samples in

individual HD centers.

Microbiological quality at dif-

ferent HD centers varied widely in treated water.

Samples showing acceptable counts ranged from

100% to a minimum of 32.8% (Tab. II). However,

water microbiological quality improved over the

study period. When results for only the last year

were assessed, the worst center had more than 60%

of acceptable samples (Tab. III).

Pseudomonas aeruginosa in water at all sampling sites. P.

aeuruginosa

was recovered from water samples from

all sampling points, except after reverse osmosis bac-

tericidal treatment (UV light or ozone), immediately

after the treated water storage tank, and from the

membrane-washing room. At the other sampling

points there was a low percentage of positive sam-

TABLE I -

HETEROTROPHIC BACTERIA COUNTS AT ALL SAMPLING POINTS

Sampling point

Number of samples

<200

200-1000

>1000

Sand filter

Activated carbon filter

596

Softener or deionizer

527

Reverse osmosis

441

Treated water storage tank

291

HD room

696

Membrane washing room

123

Recirculation loop

305

After bactericidal treatment

Total

3129

a - Percentage of samples with heterotrophic bacteria counts below 200 colony forming units (CFU)/mL

b - Percentage of samples with heterotrophic bacteria counts between 200 and 1000 CFU/mL

c - Percentage of samples with heterotrophic bacteria counts over 1000 CFU/mL

TABLE II -

HETEROTROPHIC BACTERIA COUNTS IN INDIVIDUAL HEMODIALYSIS (HD) CENTERS (JULY 1997 JUNE 2000)

HD Center

<200

200-1000

>1000

100

98.2

1.8

97.6

2.4

95.6

4.4

95.2

2.4

93.6

2.1

4.3

93.3

3.3

88.5

7.7

3.8

78.6

21.4

76.9

23.1

74.1

14.8

11.1

73.4

21.9

4.7

72.1

19.8

8.1

68.3

17.1

14.6

56.7

23.3

55.6

2.2

42.2

32.8

35.9

31.3

Total

879

a - Percentage of samples with heterotrophic bacteria counts below 200 colony forming units (CFU)/mL

b - Percentage of samples with heterotrophic bacteria counts between 200 and 1000 CFU/mL

c - Percentage of samples with heterotrophic bacteria counts over 1000 colony forming units CFU/mL

d - Number of samples

Page 4

Zunino et al

377

ples, ranging from 0.4% (one sample out of 269), at

the treated water storage tank sampling point, to

2.7% (17/632) in the HD room (Tab. IV).

Total coliforms in water from all sampling sites.

Co-

liforms were recovered from water at five sampling

points: activated carbon filter, softener or deioniser,

reverse osmosis, treated water storage tank, and HD

room. Coliforms were not recovered from the sand

filter, membrane washing room, recirculation loop,

or after bactericidal treatment (UV light or ozone)

(Tab. V). The percentage of positive samples was hi-

ghest after the carbon filter (0.8%, 4/511) and lowe-

st from the treated water storage tank (0.4%, 1/269)

(Tab. V).

TABLE III -

TOTAL HETEROTROPHIC BACTERIA COUNTS IN INDIVIDUAL HEMODIALYSIS (HD) CENTERS (JULY

1999 JUNE 2000)

HD Center

<200

200-1000

>1000

100

92.3

7.7

90.9

9.1

100

12.5

100

81.8

18.2

64.3

21.4

14.3

78.6

21.4

85.7

14.3

100

67.7

19.4

12.9

100

85.7

14.3

Total

291

a - Percentage of samples with heterotrophic bacteria counts below 200 colony forming units (CFU)/mL

b - Percentage of samples with heterotrophic bacteria counts between 200 and 1000 CFU/mL

c - Percentage of samples with heterotrophic bacteria counts over 1000 CFU/mL

d - Number of samples

TABLE IV -

PSEUDOMONAS AERUGINOSA

IN SAMPLES OBTAINED AT DIFFERENT POINTS OF THE HD SYSTEM

Sampling point

Total samples

Positive samples

Positive

Sand filter

1.1

Activated carbon filter

511

2.0

Softener or deionizer

499

1.4

Reverse osmosis

432

1.8

Treated water storage tank

269

0.4

HD room

632

2.7

Membrane washing room

210

Recirculation loop

281

1.4

After bactericidal treatment

Total

2985

1.6

a - Number of samples containing

P. aeruginosa

b - Percentage of samples containing

P. aeruginosa

Page 5

Hemodialysis water in Uruguay

378

ISCUSSION

This three-year study assessed the bacterial quality of

water used in 32 out of a total of 38 HD centers throu-

ghout Uruguay. When heterotrophic bacteria counts

were related to the different sampling points, the hi-

ghest percentage of acceptable samples (<200

CFU/mL) was obtained after UV or ozone bacterici-

dal treatment (99%). This suggests that the use of bac-

tericidal treatment devices after reverse osmosis units

in HD water purification systems has a remarkably be-

neficial effect on water bacterial quality. Water obtai-

ned after sand filters also gave a high percentage of ac-

ceptable samples (98%), suggesting that these filters

may not be critical for water bacterial quality. Howe-

ver, the number of samples obtained after UV or bac-

tericidal treatment and after the sand filter was lower

than the number of samples at other points of the sy-

stem. The lowest percentage of acceptable samples

was at the activated carbon filters (65%), softeners

and deionisers (60%) and HD rooms (76%). These re-

sults are in accordance with other reports that carbon

filters and softeners and deionisers are important risk

points since they can host bacterial growth. In addi-

tion, water going through them is already chlorine

free (7, 8).

Poor results in the HD rooms could indicate bad hy-

giene of room faucets since recirculating water quality

was notably better. However, when results for HD

room faucet samples were considered year by year, the

percentage of acceptable values rose encouragingly,

from 58% for the first year to 87% for the third. This

improvement might be due to the progressive applica-

tion of faucet disinfecting processes as routine at the

different HD centers.

Microbiological counts of treated water over the three

years showed wide variability in the different centers,

ranging from 100% (only one centre) to 32.8% of

acceptable samples. This indicates that HD centers in

Uruguay vary widely and that good practice and

methods need to be extended to all the centers.

However, looking at these measurements only for the

final year of the study, the situation had dramatically

improved. Ten centers out of 23 had 100% acceptable

samples (below 200 CFU/mL) and the worst one

showed 67.7% of acceptable samples. These results in-

dicate that, in general, control programmes have im-

proved and have had a beneficial effect on water bac-

terial quality.

The higher heterotrophic counts obtained during

warm weather suggest the need for maximum precau-

tions during this period when bacterial multiplication

is fastest. Laurence et al reported a similar correlation

between counts and warm seasons during a seven-year

multicenter study in Canada (9).

We also evaluated

Pseudomonas aeruginosa

and total co-

liforms in samples from different points of the treat-

ment systems.

P. aeruginosa

was recovered from 1.6%

of total samples. The highest percentage of positive

samples was in the HD room (2.7%), carbon filter

(2.0%), and at reverse osmosis (1.8%).

P. aeruginosa

was not detected in water from the membrane wa-

shing rooms or water obtained after UV or ozone

treatment.

Total coliforms were observed in 0.5% of samples.

The highest percentage was at the activated carbon fil-

ter (0.8%), reverse osmosis (0.7%), and softener or

deioniser and HD rooms (0.6%). Total coliforms and

Pseudomonas

recoveries were lower than those repor-

ted by Arvanitidou et al in a similar study from Greece

in 1997 (1).

In a previous survey in Uruguay, from January 1996 to

June 1997, 912 water samples from different points of

water treatment systems at 21 HD centers were analy-

sed to check the microbiological quality of water used

for HD and the efficiency of treatment (10). Only

TABLE V -

COLIFORMS IN SAMPLES OBTAINED AT DIFFERENT POINTS OF THE HD SYSTEM

Sampling point

Total samples

Positive samples

Positive

Sand filter

Activated carbon filter

511

0.8

Softener or deionizer

499

0.6

Reverse osmosis

432

0.7

Treated water storage tank

269

0.4

HD room

632

0.6

Membrane washing room

210

Recirculation loop

281

After bactericidal treatment

Total

2985

0.5

a - Number of samples containing total coliforms

b - Percentage of samples containing total coliforms

Page 6

Zunino et al

379

63.4% of the samples were under the AAMI acceptan-

ce limit. The sampling point with the highest percen-

tage of acceptable values was reverse osmosis (85.3%)

and the one with the lowest percentage was the HD

room faucets (18%) (10).

The improvement seen in the findings of our three-

year study was probably due to the adoption of better

equipment in the treatment systems at most HD cen-

ters, the addition of bactericidal treatment devices af-

ter reverse osmosis and the implementation of more

efficient water quality assurance programmes inclu-

ding regular monthly sampling and thorough disin-

fection of HD room faucets. The microbiological qua-

lity of water used for HD after the present study can be

considered among the best results reported after simi-

lar studies (1, 9, 11).

Address for correspondence:

Lucía Beltrán, M.D.

Juan Paullier 1070

11200, Montevideo, Uruguay

bzmicro@adinet.com.uy

EFERENCES

1. Arvanitidou M, Spaia S, Katsinas C, Pangidis P, Constantinidis

T, Katsouyannopoulus V, Vayonas G. Microbial quality of wa-

ter and dialysate in all haemodialysis centres of Greece.

Nephrol Dial Transplant 1998; 13: 949-54.

2. Favero MS, Alter MJ, Bland LA. Dialysisassociated infections.

In Bennett JV, Brachman PS, Sanford JP (eds.), Hospital In-

fections. Boston: Little Brown, 1992; 375-403.

3. Du Molin GC, Coleman EC, Hedley-Whyte J. Bacterial colo-

nization and endotoxin content of a new renal dialysis water

system composed of acrylonitrile butadiene styrene. Appl

Env Microbiol 1987; 53: 1322-6.

4. Morin P. Identification of the bacteriological contamination

of a water treatment line used for haemodialysis and its disin-

fection. J Hosp Infect 2000; 45: 218-24.

5. Association for the Advancement of Medical Instrumentation

(AAMI). Water quality for dialysis, 3rd. Ed., AAMI WQD, Ar-

lington, 1998.

6. American Public Health Association (APHA). Standard Meth-

ods for the Examination of Water and Wastewater, APHA-

AWWA-WEF, 20th Ed.,1998; 9-33.

7. Russell AD, Hugo WB, Ayliffe GA (eds.). Disinfection, Preser-

vation and Sterilization, Blackwell Scientific Publications

1992; 286-8.

8. Lauer J, Streifel A, Kjellstrand CM, De Roos R. The bacterial

quality of hemodialysis solution as related to several environ-

mental factors. Nephron 1975; 15: 87-97.

9. Laurence RA, Lapierre ST. Quality of haemodialysis water: a

7-year multicenter study. Am J Kidney Dis 1995; 25: 738-50.

10. Beltrán L, Zunino L, Méndez H, Martínez I, Ramos J, Grosso

R, Armand Ugón D. Contaminación microbiana de agua para

Hemodiálisis, III Jornadas Rioplatenses de Microbiología,

Buenos Aires, Argentinian Association for Microbiology, Oc-

tober 1997; 101.

11. Bambauer R, Schauer M, Jung WK, Daum V, Vienken J. Con-

tamination of dialysis water and dialysate. A survey of 30 cen-

ters. ASAIO J 1994; 40: 1012-6.

Received: January 15, 2002

Revised: April 29, 2002

Accepted: May 30, 2002