NDT Advance Access originally published online on May 17, 2007
Nephrology Dialysis Transplantation 2007 22(8):2304-2315; doi:10.1093/ndt/gfm190
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Machine-generated bicarbonate dialysate for continuous therapy: a prospective, observational cohort study
1Department of Nephrology and Hypertension, 2Quantitative Health Sciences and 3Section of Dialysis & Extracorporeal Therapy, Department of Nephrology and Hypertension, The Cleveland Clinic, Cleveland, OH 44195, USA
Correspondence and offprint requests to: Emil P. Paganini, MD, Section of Dialysis & Extracorporeal Therapy, Department of Nephrology and Hypertension, The Cleveland Clinic, Cleveland, OH 44195, USA. Email: paganie{at}ccf.org
 |
Abstract |
|---|
 Top Abstract BackgroundMethodsResultsDiscussionSupplementary materialAcknowledgementsReferences |
|---|
Background. In 1995, we described the technique of adapting a haemodialysis (HD) machine to produce a composition-adjustable, bicarbonate-based fluid (as our primary source for dialysate) for continuous HD in intensive care unit (ICU) patients with acute renal failure (ARF). The following studies the clinical effects, biochemical changes and economic costs of this practice in a large cohort of patients at a single centre over the last 10 years.
Methods. The CCF-ARF Support Registry (1995–2001) was used to identify 405 patients initially supported with bicarbonate continuous HD. The registry is a prospective, observational cohort database that captures demographic, dialysis therapy, laboratory and outcome data. All supported ARF patients were recorded from 1995–98, and then one in five patients from 1999 to 2001. We also reviewed records of the individual dialysis procedures, dialysate disposal, dialysate monitoring tests and specific costs.
Results. Continuous HD was performed for 1292 ± 587 days from 1994 to 2004. Demographics [age 59.57 ± 14.41 years, weight 84.2 ± 24 kg, male 65%, chronic kidney disease (CKD) 34%] and ICU mortality (60.5%) were comparable to other reported series. Day 4 solute [BUN 52.3 mg/dl (95% CI 49.6–54.9), creatinine 2.79 mg/dl (95% CI 2.64–2.95)], electrolyte and acid–base balance [bicarbonate 24.12 mmol/l (95% CI 23.7–24.6)] were well controlled. Dialysate monitoring revealed no positive cultures or elevated endotoxin levels. Variable-composition dialysate was achieved and delivered to all patients without adverse consequences. The cost of dialysate actually declined over time (1995 = $0.91/l, 2005 = $0.67/l).
Conclusion. We have demonstrated that ICU ARF patients can be safely, effectively and economically supported with continuous HD using this source.
Keywords: acute renal failure; bicarbonate dialysate; continuous renal replacement therapy; dialysis; mortality
|
Background |
|---|
|
TopAbstractBackgroundMethodsResultsDiscussionSupplementary materialAcknowledgementsReferences |
|---|
In 1994, the Section of Dialysis and Extracorporeal Therapy began studies into the feasibility of adapting intermittent haemodialysis (HD) machines to produce bicarbonate-based dialysate for use in continuous renal replacement therapies (CRRT), culminating in consistent production of large quantities of clean, safe and composition-variable dialysate [1]. Bicarbonate-based dialysate is believed to be more physiological and efficient in correcting acid–base disturbances in acute renal failure (ARF) patients, and avoids some of the disadvantages of lactate-based fluids such as, hypotensive episodes, reduced mean arterial pressures, increased urea generation, cerebral dysfunction or hyperlactataemia with liver dysfunction [2]. Since then, we have almost exclusively supported our ARF patients in the intensive care units (ICU) with this dialysate on continuous veno-venous haemodialysis (CVVHD).
Briefly, the dialysate is prepared using a volumetric-controlled, single-pass dialysis machine (Althin/Baxter-1000; Baxter, Deerfield, IL) [1]. The machine proportions and mixes two concentrates (acid and bicarbonate) with heated ultra-pure water (produced by reverse osmosis in our dialysis unit) into the dialysate with the required composition. This dialysate is inflowed at 500–900 ml/min into the dialysate compartment of a high-flux, hollow-fibre, polysulphone membrane dialyser (currently, Fresenius F160; Fresenius AG, Bad Homburg, Germany), and allowed to transfer by back-filtration from the dialysate compartment into the dialyser ‘blood’ compartment, from which it is drained into 6- to 8-l bags (Fresenius). This method has been essentially unchanged for the last 10 years, except for the consumables used, and the protocols instituted for safety monitoring (Figure 1).
|
The CRRT literature is replete with many descriptions and studies of using different techniques for clearance [34]. Methods using HD rely on diffusion for clearance; haemofiltration by convection, or a combination of both, haemodiafiltration [5]. CVVHD is used primarily in our institution because the sterility of the machine-generated dialysate cannot be assured, so direct infusion for haemofiltration is not recommended [6]. We decided to study the clinical effects, biochemical changes and economic costs of the practice of continuous HD in a large cohort of patients with this technique of bicarbonate dialysate generation over the last 10 years.
|
Methods |
|---|
|
TopAbstractBackgroundMethodsResultsDiscussionSupplementary materialAcknowledgementsReferences |
|---|
This study was approved by the Institutional Review Board. For clinical data, we abstracted the ARF support registry from 1995 to 2001. This registry is a prospective, observational cohort study that captured demographic, dialysis therapy, laboratory and outcome data on ARF patients in the ICU who required support. All ARF patients were recorded in the registry from 1995 to the first quarter of 1998, and then a randomly selected one in five patients were recorded for the remainder of 1998–2001. A biostatistician used a computer program to randomly generate the number of patients between recruitment, while keeping the overall ratio of 1 : 5. For this study, patients who were supported with machine-generated bicarbonate dialysate CRRT as the initial therapy were included. The longer-term survival of these patients was determined by matching their social security numbers, when available, to the social security death index (updated till 2004). The volume of dialysis was calculated from the total effluent discharged per day divided by the patient's baseline weight or pre-morbid weight; and expressed as millilitre per kilogram per hour, thus capturing the periods of non-dialysis during interruption and incorporating that downtime into an average of 24 h exposure to dialysate volume.
We reviewed CRRT procedure logs from 1994 to 2004 for the quantity and types of therapies performed. The random monitoring tests for the dialysate, records of discarded dialysate and cost data were also analysed. The random monitoring tests for the dialysate comprise of Limulus amoebocyte lysate (LAL) assays for bacterial endotoxin. By protocol, two bags are randomly selected per month and samples drawn for LAL assays. The bags are isolated until the assay results are reported and documented. If the LAL assay is 
1 EU/ml (endotoxin unit per millilitre) the test is immediately repeated. If results are <1 EU/ml, the bags will be discarded. The results were trended and reported to the medical director each quarter. Longer-term testing of dialysate consisting of bacterial cultures, chemical analyses and endotoxin levels were described previously and will not be discussed further in this study [1,6]. The choice of the LAL assay and the test threshold was chosen to meet or exceed the standards set by the American Association of Medical Instrumentation (AAMI) for endotoxin levels in water and concentrates required for producing dialysate used in HD.
Statistical analysis
Continuous baseline measures were described using mean ± standard deviation (SD) or with 95% confidence intervals; or median with inter-quartile range (IQR). Categorical baseline measures were described using frequencies and percentages. The average volume of dialysis in the first 3 days of therapy was calculated only for those patients with complete information for all 3 days. Differences in mean volume by gender, and by years, 1995–1998 vs 1999–2001, were assessed by using two sample t-tests for unequal variance. A repeated measures mixed model was used to assess the association between nutritional supplementation and blood urea nitrogen (BUN) levels over time, while adjusting for baseline BUN. A within-subject spatial power correlation structure was chosen to allow for the correlation between measurements to decrease by the distance between time points. The Kaplan–Meier method was used to estimate survival time by gender, and change to alternative therapy. The groups were compared on survival time using the log rank test. ICU mortality was computed as the time from ICU admission to death, patients who were alive at the time of ICU discharge were censored, and no patients were lost to follow-up. Hospital mortality was computed as the time from hospital admission to death, patients who were alive at the time of hospital discharge were censored, and no patients were lost to follow-up. Cox regression was used to assess the relationship between each baseline measure of interest with time to death, while adjusting for the volume of dialysis (main variable of interest). Out of 405 patients, volume of dialysis was calculated for 299 patients who had complete therapy information for the first three days of therapy. If albumin was missing at baseline, the first value recorded within 3 days of the start of therapy was used. Baseline measures used included age, gender, presence of respiratory failure, white cell count, haematocrit, platelets, pH, CO2, BUN, creatinine, glucose, albumin, total bilirubin, serum aspartate aminotransferase (AST) and the APACHE II score at the time of renal consult. Baseline measures at the 0.10 significance level were kept, and a complete case analysis for each outcome was done using backward selection. All testing was done at the 5% significance level.
|
Results |
|---|
|
TopAbstractBackgroundMethodsResultsDiscussionSupplementary materialAcknowledgementsReferences |
|---|
From 1994 to 2004, we performed an average of 1551 ± 530 days of therapy per year of CRRT. Of these, 1492 ± 512 therapy days were continuous HD (Table 1). Therapy with continuous veno-venous haemodiafiltration (CVVHDF) increased substantially in 2004 due to our participation in the NIH/VA ATN Study [7]. Conversely, continuous arterio-venous haemofiltration (CAVH) and CVVH decreased substantially, being applied primarily for therapeutic clearance of high molecular weight substances in patients with toxic overdose or poisonings. The number of patient-therapy days with veno-venous access increased from 660 in 1994 to 2246 in 2004, and therapy with arterio-venous access declined from 126 to 34.
|
The observational cohort study included 405 patients who developed ARF in the ICU and were started on continuous arterio-venous haemodialysis (CAVHD) or CVVHD as the initial therapy. For all years, the baseline characteristics of the patients and their therapies are shown in Tables 2 and 3. Sixty-five percent of the patients were male, and 34% had chronic kidney disease [defined as unstressed serum creatinine >1.6 mg/dl (144 µmol/l) on admission to the hospital]. The patients had, on average, significant azotaemia (BUN = 83 ± 41.7 mg/dl, or 29.6 ± 14.9 mmol/l), hyperglycaemia (177 ± 92 mg/dl, or 9.73 ± 5.1 mmol/l), hyperphosphataemia (5.8 ± 2.3 mg/dl, or 1.87 ± 0.74 mmol/l) and leucocytosis (15.7 ± 8.7 x 103 counts/mm3). Of all patients, 88.6% were started on CVVHD and 32.8% changed to intermittent therapies during their ICU course; only 15.1% of patients received heparin anticoagulation during the first day of therapy.
|
|
Volume of dialysis
On average, CRRT was started 4 days (IQR: 2–10) after admission to the ICU. The median duration of CRRT, until therapy was stopped or changed, was 5 days (IQR: 3–9). The overall mean volume of dialysis (ml/kg/h) on day 2 was 14.8 (95% CI 14.1–15.5). Male patients received less therapy than females, 14.0 (95% CI 13.2–14.8) and 16.3 (95% CI 15.0–17.7), respectively (P = 0.005). The overall first 3-day average dialysis volume was 13.2 (95% CI 12.6–13.8). This was also higher in female patients than males, 14.5 (13.3–15.6) and 12.5 (95% CI 11.9–13.1), respectively (P = 0.003). The mean volume of dialysis was also higher in years 1999–2001 as a group than years 1995–1998 as a group, 15.8 (95% CI 14.3–17.4) vs 12.6 (95% 12.0–13.1, P < 0.001), respectively (Figure 2). The longer the duration of therapy received, the higher the average volume of dialysis delivered, achieving >14 ml/kg/h of dialysis from Day 6 onwards (Figure 3).
|
|
Clearance and electrolytes
Serum electrolytes, glucose, bicarbonate, BUN and creatinine were well maintained during therapy (Figure 4). After 3 days of therapy, bicarbonate levels (mmol/l) rose from 19.85 (95% CI 19.3–20.4) at baseline to plateau at 24.08 (95% CI 23.63–24.54). After 5 days of therapy, BUN increased even though dialysis dosing also increased. This appears to be related to more patients placed on nutritional support over time (Figure 5). Patients on nutritional supplementation had lower BUN levels in the first few days of therapy, and higher BUN levels during Days 3 through 12, however, these differences were not significant (P = 0.19).
|
|
Mortality
The ICU mortality was 60.5% and hospital mortality was 71.4%, which are similar to previous reports of ARF patients requiring RRT (Table 4); the mortality was directly comparable for US patients in the observational study of PICARD, but other quoted statistics are not directly comparable since the studies differed in patient populations, types of RRT, the setting of therapy delivery and the personnel involved in decision-making and therapy delivery [8–10]. The volume of dialysis was not associated with ICU mortality [hazard ratio (HR) = 1.09, 95% CI 0.82–1.43, P = 0.56] (Table 5, Figure 6), nor with in-hospital mortality (HR = 0.80, 95% CI 0.50–1.29, P = 0.36). Higher baseline BUN and CO2 levels were, however, consistently associated with lower ICU and in-hospital mortality rates. The median length of stay of patients who died in the hospital was 20 days (IQR = 11–34), and the median length of stay of survivors was 46.5 days (IQR = 25–73). Gender did not impact the probability of survival (log rank P = 0.95) (Figure 7). However, a change from CRRT to intermittent therapy was associated with an increase in the probability of survival (log rank P < 0.001) (Figure 8).
|
|
|
|
|
Dialysate safety and cost
All formulations of dialysate were delivered to the patient's bedside in
30 min from the time of prescription. There were no positive results for the 3 years of dialysate monitoring tests (LAL assays). There were also no reported incidences of complications or adverse effects with the dialysate or the production method. Information on discarded dialysate bags was recorded intermittently from 2002 to 2004 (21 months). The predominant reason for disposal of dialysate was expiration after 72 h from production (Table 6). For 11 months in 2005, we used 494 cases of bags (30 bags per case) for dialysate production. Therefore, we discarded an average of 55 bags ± 29 bags per month, or about 4% of total production per month. The cost per litre of dialysate produced decreased from $0.91 per litre in 1995 to $0.67 per litre in 2005 (Table 7). This is mainly due to changes in the costs of materials. However, the cost of labour increased from $15.30 per hour to $19.01 per hour. This represents 46% of total costs in 2005 as compared to 1995, where it was only 28%. We had changed from 12–15 l bags (Cycler drainage set; Baxter Healthcare Corporation, Deerfield, IL) to 6–8 l bags (Fresenius), and from F80 to Optiflux 160 dialyser (Fresenius). These two items are less expensive now than 10 years ago.
|
|
|
Discussion |
|---|
|
TopAbstractBackgroundMethodsResultsDiscussionSupplementary materialAcknowledgementsReferences |
|---|
In terms of safety, efficacy, economic costs and efficiency, this study provides the long-term clinical validation of using large quantities of bicarbonate dialysate produced by the machine method in a large cohort of patients treated with continuous HD. The machine method has the advantage of rapidly producing large quantities of composition-variable ultrapure dialysate economically for use in CRRT at the doses of dialysis employed currently. The higher volumes of dialysis from 1999 can be attributed to a significant practice change in our institution when it was observed that ARF patients in the ICU had increased survival with higher doses of dialysis. This observation was noted initially by our group with retrospective data from the ARF support registry [11]. A subsequent randomized controlled trial further supported the use of a higher volume of haemofiltration in critically ill patients on CRRT [12]. In our current study, we have not been able to show a mortality benefit with the volume of dialysis. This may be for several reasons. In the oft-quoted study on the volume of haemofiltration in CRRT (haemofiltration), the dose that demonstrated a survival advantage was 35 ml/kg/h [12]. The lower range in the volumes of dialysis achieved in our cohort undermined our statistical power to detect a survival advantage. Secondly, there may be a difference to using haemofiltration vs HD in critically ill patients (discussed further). Finally, even if associations are determined from such an analysis, the significance of these findings is limited by the observational nature of the data. Our current practice is to prescribe a minimum of 35 ml/kg/h of continuous HD. Assuming that the average 84 kg patient requires 35 ml/kg/h of CRRT dose, he will need 2940 ml/h of dialysate or haemofiltration (about 3 l/h). The dialysate will thus cost $2.01 per hour (or $48.24 per day). In our institution, the acquisition cost of commercially packaged dialysate or replacement fluids is $22 per 5 l bag, or $4.40 per litre (Prismasate; Gambro Renal Products, USA, Lakewood, CO). If the patient received this fluid, it will cost $13.20 per hour (or $316.80 per day). In contrast, should the patient require customized formulations, the fluid produced by a compounding pharmacy will approach $14 per litre (or $1008 per day). Thus, given the economic costs associated with the higher doses of CRRT being offered today, it may be prohibitive to support patients for long periods with commercially, or pharmacy-prepared fluids in haemofiltration or haemodiafiltration. This method is particularly attractive for curtailing costs in high-dose CVVHD programs while we await the arrival of machines capable of on-line generation of dialysate.
Because of the higher costs associated with CRRT and no clear evidence of therapeutic superiority over intermittent HD, hybrid techniques using conventional intermittent HD machines have been developed. These techniques, variously labelled as sustained low-efficiency dialysis (SLED) or extended daily dialysis (EDD), attempt to offer the advantages of CRRT of slower removal of large volumes of fluid and higher total clearance in haemodynamically compromised, volume-overload and catabolic patients, and yet contain the costs of delivering therapy [13]. Our technique of delivering CRRT is similar to these methods in that we use these machines to generate the dialysate used, except that the additional filtration of dialysate is not available in SLED. The main difference lies in the total time of therapy delivered, although this may be more perceived than real. In practice, continuous therapy is rarely continuous as it is interrupted by failure of circuit circulation due to clotting or vascular access flow disturbances, and temporary disruptions for diagnostic or therapeutic procedures. Additionally, SLED has evolved to the point that it may be indistinguishable from CRRT [14].
In addition to cost, another major advantage of the technique is the ability to rapidly produce large volumes of customized bicarbonate solutions with varying concentrations of electrolytes. Compounding these solutions in a pharmacy is costly, time-consuming and may be prone to errors [15]. The machine technique offers the advantage of built-in safety checks when the conductivity of the final solution is continuously monitored during rapid proportioning of the component solutions. Potassium, calcium, bicarbonate and sodium levels may be adjusted to suit the clinical requirements of the patient. For example, we routinely produce solutions with no calcium, and lower sodium and bicarbonate concentrations for use in patients who need to be placed on regional citrate anticoagulation. One of the shortcomings of CRRT is the inability to flexibly alter the amount of buffer and electrolytes delivered. Machine-generated dialysate overcomes this limitation. As our understanding of acid-base disorders in critically ill patients develops further with the physico-chemical analyses of strong ions in the serum, it is also recognized that the relative proportions of strong ions in solution affects the pH of that solution, a potentially important observation since the ability to adjust the constituent electrolytes of a dialysate would then change its pH [16]. In practice, we noted that patients were developing metabolic alkalosis sooner as the volume of dialysis increased over the years. As such, the bicarbonate concentration was adjusted downwards. From 1994 to 1996, the bicarbonate concentration was 39 mmol/l; from 1997 to 2003, it was 35 mmol/l and from 2004 onwards, it was further reduced to 32 mmol/l. Therefore, the ability to rapidly produce large volumes of customized bicarbonate solutions is a distinct advantage in managing complex and critically ill patients with highly variable metabolic and electrolyte abnormalities.
A major limitation of the machine method is the inability to use the machine-generated dialysate for haemofiltration or haemodiafiltration due to higher sterility requirements, therefore, we have almost exclusively supported our patients with continuous haemodialysis [1,6]. We had previously reported that CRRT circuits and prepared dialysate bags may become contaminated with bacterial growth, hence, monitoring and practice protocols were instituted to ensure safety [6]. Currently, two bags per month are randomly selected and tested for bacterial endotoxin, and bags of dialysate not used within 72 h from the time of production are discarded. In addition, our CRRT circuits undergo a complete change of tubing sets and dialysers every 72 h. These protocols have ensured that in the 10 years of this practice, we have not recorded any adverse event or complication associated with the dialysate; although this is limited by having no standard classification of adverse events or complications. (E.P. Pagnini, personal communication. At the attending nephrologist's discretion, severe derangements in electrolytes during CRRT, or non-correction of serum abnormalities despite adequate CRRT, prompted chemical analyses of the delivered dialysate, and these were reported to be within production specifications.)
Arguably, the largest clinical trial showing a survival benefit with the dose of CRRT in ARF patients was performed with CVVH [12]. However, there are no direct comparisons of CVVH/CVVHDF with CVVHD that show superiority in terms of survival between these different techniques of CRRT. While small molecule clearance should be equal (with equal flow rates up to a point) in HD, middle molecule clearance may be better with haemofiltration or combined techniques [5]. But whether additional clearance of these middle molecules has any impact on mortality is also unknown. Moreover, current research suggests that advances in membrane technology may reduce or obviate the advantages of convective over diffusive clearances of middle molecules, especially with the long therapy times associated with CRRT [17]. Additionally, calcium-containing bicarbonate solutions can be susceptible to crystallization [18]. While clinical experience has not detected substantial deleterious effects with infusing these solutions as a replacement fluid during haemofiltration, dialyser membranes are an added barrier to these microcrystals during HD.
The higher concentration of glucose, 205 ± 15 mg/dl (11.14 ± 0.83 mmol/l) in the dialysate produced may be disadvantageous [1]. Insulin resistance, defined as hyperglycaemia in the setting of hyperinsulinaemia, occurring in patients with ARF is associated with increased mortality [19]. The mean blood glucose levels attained in our patients while on dialysis were well above 110 mg/dl (6.1 mmol/l) (Figure 4H). In critically ill patients, insulin therapy to control blood glucose below 110 mg/dl (6.1 mmol/l) has been shown to improve overall mortality and morbidity, and also reduces the incidence of ARF requiring dialysis [20]. This suggests that lowering the glucose level in dialysate preparations, as well as initiating insulin therapy for these patients may be required.
In summary, we have shown that it is technically possible to generate ultrapure dialysate by the machine method and provide CVVHD to a large number of critically ill patients with ARF. The dialysate is efficacious in controlling electrolyte derangements and uraemia; and the production method has the added flexibility of safely adjusting the composition of the dialysate, while generating large quantities of dialysate quickly at substantially lower costs. Additionally, the dialysate maintained its efficacy and stability with the packaging and storage conditions described, and therapy was delivered with a high degree of safety. In continuous forms of renal support therapy, there is no proven advantage of convection over diffusion because of the effect of time on middle molecule clearances.
In conclusion, we have therefore demonstrated that a large number of ARF patients in the ICU can be safely, effectively and economically supported with continuous HD using machine-generated bicarbonate-based dialysate, customized for electrolyte composition.
|
Supplementary material |
|---|
|
TopAbstractBackgroundMethodsResultsDiscussionSupplementary materialAcknowledgementsReferences |
|---|
For Supplementary Material, please refer to NDT Online.
|
Acknowledgements |
|---|
|
TopAbstractBackgroundMethodsResultsDiscussionSupplementary materialAcknowledgementsReferences |
|---|
The authors wish to acknowledge the contribution of Tracy Siefert, RN in the maintenance of the database. This study was presented in part at the 11th Annual International Conference on Continuous Therapies (March 2006) at Coronado, CA, USA.
Dr Teo is a Clinical Research Scholar supported by the Department of Nephrology and Hypertension, The Cleveland Clinic.
Conflict of interest statement. None declared.
|
Notes |
|---|
The authors wish it to be known that, in their opinion, the first two authors contributed equally to this work.
|
References |
|---|
|
TopAbstractBackgroundMethodsResultsDiscussionSupplementary materialAcknowledgementsReferences |
|---|
- Leblanc M, Moreno L, Robinson OP, Tapolyai M, Paganini EP. Bicarbonate dialysate for continuous renal replacement therapy in intensive care unit patients with acute renal failure. Am J Kidney Dis (1995) 26:910–917.[Web of Science][Medline]
- Mehta RL. Acid-base and electrolyte management in continuous renal replacement therapy. Blood Purif (2002) 20:262–268.[CrossRef][Web of Science][Medline]
- Brause M, Neumann A, Schumacher T, Grabensee B, Heering P. Effect of filtration volume of continuous venovenous hemofiltration in the treatment of patients with acute renal failure in intensive care units. Crit Care Med (2003) 31:841–846.[CrossRef][Web of Science][Medline]
- Page B, Vieillard-Baron A, Chergui K, et al. Early veno-venous haemodiafiltration for sepsis-related multiple organ failure. Crit Care (2005) 9:R755–R763.[CrossRef][Web of Science][Medline]
- Parakininkas D, Greenbaum LA. Comparison of solute clearance in three modes of continuous renal replacement therapy. Pediatr Crit Care Med (2004) 5:269–274.[CrossRef][Medline]
- Kanagasundaram NS, Larive AB, Paganini EP. A preliminary survey of bacterial contamination of the dialysate circuit in continuous veno-venous hemodialysis. Clin Nephrol (2003) 59:47–55.[Web of Science][Medline]
- Palevsky PM, O’Connor T, Zhang JH, Star RA, Smith MW. Design of the VA/NIH Acute Renal Failure Trial Network (ATN) Study: intensive versus conventional renal support in acute renal failure. Clin Trials (2005) 2:423–435.
[Abstract/Free Full Text] - Metnitz PG, Krenn CG, Steltzer H, et al. Effect of acute renal failure requiring renal replacement therapy on outcome in critically ill patients. Crit Care Med (2002) 30:2051–2058.[CrossRef][Web of Science][Medline]
- Morgera S, Kraft AK, Siebert G, Luft FC, Neumayer HH. Long-term outcomes in acute renal failure patients treated with continuous renal replacement therapies. Am J Kidney Dis (2002) 40:275–279.[CrossRef][Web of Science][Medline]
- Mehta RL, Pascual MT, Soroko S, et al. Spectrum of acute renal failure in the intensive care unit: the PICARD experience. Kidney Int (2004) 66:1613–1621.[CrossRef][Web of Science][Medline]
- Paganini EP, Tapolyai M, Goormastic M. Establishing a dialysis therapy/patient outcome link in intensive care unit acute dialysis for patients with acute renal failure. Am J Kidney Dis (1996) 28:S81–S89.[Web of Science]
- Ronco C, Bellomo R, Homel P, et al. Effects of different doses in continuous veno-venous haemofiltration on outcomes of acute renal failure: a prospective randomised trial. Lancet (2000) 356:26–30.[CrossRef][Web of Science][Medline]
- Berbece AN, Richardson RM. Sustained low-efficiency dialysis in the ICU: cost, anticoagulation, and solute removal. Kidney Int (2006) 70:963–968.[CrossRef][Web of Science][Medline]
- Finkel KW, Foringer JR. Safety of regional citrate anticoagulation for continuous sustained low efficiency dialysis (C-SLED) in critically ill patients. Ren Fail (2005) 27:541–545.[CrossRef][Web of Science][Medline]
- Johnston RV, Boiteau P, Charlebois K, Long S. Responding to tragic error: lessons from Foothills Medical Centre. Cmaj (2004) 170:1659–1660.
[Free Full Text] - Kellum JA. Determinants of plasma acid-base balance. Crit Care Clin (2005) 21:329–346.[CrossRef][Web of Science][Medline]
- Leypoldt JK, Murray PT, Kamerath CD, Gilson JF, Friederichs G. Improved Diffusive Solute Clearances with Novel CRRT Equipment and Dialyzer Design. Blood Purif (2006) 24:247–273.[CrossRef]
- Maccariello E, Rocha E, Dalboni MA, Ferreira AT, Draibe S, Cendoroglo M. Customized bicarbonate buffered dialysate and replacement solutions for continuous renal replacement therapies: effect of crystallization on the measured levels of electrolytes and buffer. Artif Organs (2001) 25:870–875.[CrossRef][Web of Science][Medline]
- Basi S, Pupim LB, Simmons EM, et al. Insulin resistance in critically ill patients with acute renal failure. Am J Physiol Renal Physiol (2005) 289:F259–F264.
[Abstract/Free Full Text] - van den Berghe G, Wouters P, Weekers F, et al. Intensive insulin therapy in the critically ill patients. N Engl J Med (2001) 345:1359–1367.
[Abstract/Free Full Text]
Accepted in revised form: 12. 3.07
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||