Introduction
Patients on maintenance haemodialysis or haemofiltration are exposed to 20-30,000 litres of water annually. To minimise the small but defined risk of acute and chronic complications related to chemical imbalance or microbial contamination of water and dialysis systems of HD machines,- guidelines have been regularly published by the AAMI (revised 1993), the KHA/ANZSN D&T S/C (revised 1993) and in the European
Pharmacopoeia (revised 1992).
The supplier of the dialysate supply system is responsible for recommending a system capable of meeting the following guidelines for chemical and microbial purity. The dialysis physician bears the ultimate responsibility for procedures and monitoring practices that assure the safety and quality of water used for dialysis.
Chemical Purity
Risks from chemical contamination of water for dialysis include, among others, acute haemolysis from chloramines, and a cerebral-musculoskeletal syndrome from chronic aluminium exposure. These risks are significantly reduced by water chemical analysis prior to installation of dialysis, the use of a water softener and carbon filter, and the use of reverse osmosis machines. R.0. membranes can rupture or become saturated, (e.g. by clay suspension in local water), and expire prior to their expected life of 2 years. Monitors of total conductivity of final dialysate can become unstable in conditions of extreme heat or cold.
Microbiological Purity
The frequency of pyrogen reactions has been estimated in the USA at 1-5% of all dialyses, falling to 0.03-0.07% in units with in-line dialysate filtration, even in the presence of upstream contamination of water/dialysate systems. The main theoretical purpose of disinfection of the water/dialysate pathways is to minimise the frequency of pyrogen reactions from water-borne bacterial endotoxin or fragments. This is particularly important for haemofiltration, where filtration fluid derived from local water supply may be infused directly into the patients' blood stream (via a 0.2 um submicron filter). However, the clinical contribution of maintaining microbial water purity to the development of long-term complications such as amyloidosis remains uncertain.
There is little or no risk of infection or cross-infection of conventional haemo-dialysis patients from the water and dialysate pathways of a HD machine. Although liquid bicarbonate dialysate concentrate can support rapid bacterial proliferation, (i.e. 2-4 logs growth within 72 hours after opening), bacteria, viruses, spores, algae and fungi cannot cross an intact dialysis membrane (even high-flux) due to their size. Except in specific circumstances (see below), the flow of dialysate pathways is unidirectional or single-pass, and bloodborne agents contaminating the dialysate are lost as effluent.
Although blood-to-blood transmission of infectious agents is a major risk in dialysis units, transmission of blood-borne organisms from an infectious patient via the dialysate water systems to a subsequent patient could only occur in the following theoretical sequence of technical and procedural failures:
Febrile (pyrogen) reactions in patients on HD have therefore been attributed by some authors to either back-filtration or back-diffusion into the patient of dialysable bacterial endotoxin or its fragments within dialysate (or possibly from blood leukocyte pseudopodia interacting with endotoxin bound to the dialysate side of the membrane).
Endotoxins produced by naturally-occurring water organisms (e.g. Pseudomonas or non-tuberculous Mycobacteria species) are 1 -1 log less potent cytokine activators than endotoxins derived from -coliforms,- and cause no cytokine activation at clinically-relevant concentrations. Assays for water-borne bacteria may be better-performed on low-nutrient media, and cytokine activation detected with mononuclear-cell assay rather than LAL assay, although not widely available.
Disinfection of water/dialysate pathways can be performed by formaldehyde (but with irritating vapours), hypochlorite (but corrosive, -and then rinsed with potentially -contaminated dialysate water before use), peracetic acid, glutaraldehyde/hydrogen peroxide or heated water with varying degrees of completeness of sterility, time required and cost.
There is no consistent standard for recommended frequency of machine disinfection: current AAMI documentation describes daily disinfection of machines as common practice, attending dialysate systems weekly and reverse osmosis units monthly. More frequent disinfection, e.g. between every patient treatment, poses a significant extra cost in staff time and consumables, for no objective gain.
Recommendations
1. DESIGN
1. 1 Prior to installation of a HD system, local water should be investigated by enquiry with the local water supply authority for normal quality and seasonal variation.
1.2 Water samples must be analysed for the following chemicals: calcium, magnesium, sodium, potassium, fluoride, free chlorine and chloramine. Depending on local conditions, water samples may also need analysis for the following chemicals: nitrate, sulphate, copper, barium, zinc, aluminium, arsenic, lead, silver, cadmium, chromium, selenium, mercury and others as required
1.3 Maximum allowable concentrations in final feed water to the dialysis machine should not exceed the concentrations defined in the-AAMI Guidelines 1993 (attached).
1.4 Optimum water treatment systems may include particulate filtration, carbon filtration, water softening, reverse osmosis and de-ionisation. Depending on local conditions, they may also include water storage tanks, repressurisation pumps, chlorine dosing, or water coolers.
1. 5 Water/dialysate delivery systems should be designed to avoid blind loops or areas inaccessible to disinfectant
2. PRACTICE
2.1 Liquid bicarbonate dialysate concentrate should not be used more than 24 hours after opening.
2.2 Water/dialysate pathways should be disinfected using an approved system (as above) with adequate contact time. Disinfection should be routinely performed at least daily, preferably after the longest period of machine inactivity. Disinfection must be performed after any episodes of dialyser fibre rupture and blood leak, prior to the machine being used for another patient Disinfection after usage by patients with identified blood-borne infections may be performed at individual units' discretion.
3. MONITORING
3.1 Chemical analysis of the final dialysate mixture should be continuously monitored during dialysis by total conductivity, with a minimum accuracy of ± 0.1 mS of maximum indication, and temperature compensation.
3.2 Repeat water and final dialysate chemical analysis should be performed one month after installation, then at least three-monthly intervals thereafter (at least annually for Home HD patients) with more frequent monitoring in the presence of unsatisfactory results
3.3. Reverse osmosis unit membranes should be tested for percentage rejection, and exchanged if this falls below 75%. This should occur at installation, and at six-monthly intervals, or if dialysate chemical analyses are unsatisfactory.
3.4 Samples of local water supply, post-filter water, pre-machine proportioner and post-dialyser dialysate should be cultured for bacterial contamination at initial installation of the dialysis system and thereafter, using the following technique:
i) samples assayed within 30 minutes at room temperature, or 24 hours at 4ºC
ii) Total viable colony-forming units (counted after 48 hrs at 37ºC on tryptic soy agar media) to be less than 200 CFUM1 in water pre-proportioner, and less than 2000 CFUM1 in post-dialyser dialysate.
3.5 Analyses for endotoxin concentrations are not routinely required.
Originally prepared as i) Recommendations for Chemical Purity of Water for Dialysis by …. in …., revised in …. by …., and in 1993 by …., and ii) Recommendations for Microbiological Purity of Water for Dialysis by …. in …., revised in …. by …., and in 1993 by …. Amalgamated and revised by M. Thornas, F. Ordynski and RPH Infection Control Team in 1996.
References
Association for the Advancement of Medical Instrumentation. Haemodialysis systems. Arlington, VA; AAMI, 1995. American National Standard, pp-257-277.
Bland LA, Ridgeway MR, Aguero SM, Carson LA, Favero MS: Potential bacteriologic and endotoxin hazards associated with liquid bicarbonate concentrate. ASAIOTrans 1987, 33:542-545
European Pharmacopoeia 1992.
de Felice A, Cappelli G, Facchini F, Tetta C, Comia F, Airno- G, Lusvarghi E. Ultrafiltration and endotoxin removal from dialysis fluids. Kidney Int 43, SuppI 41(1993):S201-204.
Kumano K, Yokota S, Nanbu M, Sakai T. Do cytokine-inducing substances penetrate through dialysis membranes and stimulate monocytes?. Kidney Int 43, Suppi 41(1993):S205-208.
Lonnernan G. Dialysate bacteriological quality and the permeability of dialyser membranes to pyrogens. Kidney Int 43:(Suppi 4 1), S195-200, 1993.
Oettinger CW, Arduino MJ, Oliver JC, Bland LA. The clinical relevance of dialysate sterility. Seminars in Dialysis 1994, 7(4)163-267.
Powell AC, Bland LA- Oettinger CW,McAllister SK, Oliver JC, Arduino MJ, Favero MS. Lack of plasma interieukin-1 or tumor necrosis factor- elevation during unfavourable hemodialysis conditions. J Am Soc Nephrol 1991;2:1007-1013.
Tielemin C, Husson C. Schurmans T, Gastaidello K., Madhoun P. Delville J-P, Marchant A, Goldman M. Vanherweghern J-L. Effects of ultrapure and non-sterile dialysate on the inflammatory response during in vitro hemodialysis. Kidney Int 49 (1996)236-243.
