Costs and length of stay associated with antimicrobial resistance in acute kidney injury patients with bloodstream infection.

INTRODUCTION: Antimicrobial resistance negatively impacts on prognosis. Intensive care unit (ICU) patients, and particularly those with acute kidney injury (AKI), are at high risk for developing nosocomial bloodstream infections (BSI) due to multi-drug-resistant strains. Economic implications in terms of costs and length of stay (LOS) attributable to antimicrobial resistance are underevaluated. This study aimed to assess whether microbial susceptibility patterns affect costs and LOS in a well-defined cohort of ICU patients with AKI undergoing renal replacement therapy (RRT) who developed nosocomial BSI. METHODS: Historical study (1995-2004) enrolling all adult RRT-dependent ICU patients with AKI and nosocomial BSI. Costs were considered as invoiced in the Belgian reimbursement system, and LOS was used as a surrogate marker for hospital resource allocation. RESULTS: Of the 1330 patients with AKI undergoing RRT, 92 had microbiologic evidence of nosocomial BSI (57/92, 62% due to a multi-drug-resistant microorganism). Main patient characteristics were equal in both groups. As compared to patients with antimicro-4 bial-susceptible BSI, patients with antimicrobial-resistant BSI were more likely to acquire Gram-positive infection (72.6% vs 25.5%, P<0.001). No differences were found neither in LOS (ICU before BSI, ICU, hospital before BSI, hospital, hospital after BSI, and time on RRT; all P>0.05) or hospital costs (all P>0.05) when comparing patients with antimicrobial-resistant vs antimicrobial-susceptible BSI. However, although not statistically significant, patients with BSI caused by resistant Gram-negative-, Candida-, or anaerobic bacteria incurred substantial higher costs than those without. CONCLUSION: In a cohort of ICU patients with AKI and nosocomial BSI undergoing RRT, patients with antimicrobial-resistant vs antimicrobial-susceptible Gram-positive BSI did not have longer hospital stays, or higher hospital costs. Patients with resistant "other" (i.e. Gram-negative, Candida, or anaerobic) BSI were found to have a distinct trend towards increased resources use as compared to patients with susceptible "other" BSI, respectively.

Uremic toxins in chronic renal failure.

The uremic syndrome is a complex mixture of organ dysfunctions, which is attributed to the retention of a myriad of compounds that under normal conditions are excreted by healthy kidneys. During recent years major steps have been taken in the area of identification and characterization of uremic retention solutes and in the knowledge of their pathophysiological importance; however, our knowledge remains far from complete. In the present paper the general classification based on their molecular weight and on their protein-binding characteristics, with reflections on their removal, will be discussed. In addition, current knowledge about the main uremic retention products and their clinical and biological effects will be reviewed in detail.

Cytokine removal during continuous hemofiltration in septic patients.

A potential application of the continuous renal replacement therapies is the extracorporeal removal of inflammatory mediators in septic patients. Cytokine elimination with continuous renal replacement therapies has been demonstrated in several clinical studies, but so far without important effects on their serum concentrations. Improved knowledge of the cytokine removal mechanisms could lead to the development of more efficient treatment strategies. In the present study, 15 patients with septic shock and acute renal failure were observed during the first 24 h of treatment with continuous venovenous hemofiltration (CVVH) with an AN69 membrane. After 12 h, the hemofilter was replaced and the blood flow rate (QB) was switched from 100 ml/min to 200 ml/min or vice versa. Pre- and postfilter plasma and ultrafiltrate concentrations of selected inflammatory and anti-inflammatory cytokines were measured at several time points allowing the calculation of a mass balance. Cytokine removal was highest 1 h after the start of CVVH and after the change of the membrane (ranging from 25 to 43% of the prefilter amount), corresponding with a significant fall in the serum concentration of all cytokines. The inhibitors of inflammation were removed to the same extent as the inflammatory cytokines. Adsorption to the AN69 membrane appeared to be the main clearance mechanism, being most pronounced immediately after installation of a new membrane and decreasing steadily thereafter, indicating rapid saturation of the membrane. A QB of 200 ml/min was associated with a 75% increase of the ultrafiltration rate and a significantly higher convective elimination and membrane adsorption than at a QB of 100 ml/min. The results indicate that optimal cytokine removal with CVVH with an AN69 membrane could be achieved with a combination of a high QB/ultrafiltration rate and frequent membrane changes.

Pretransplantation hemodialysis strategy influences early renal graft function.

The influence of the pretransplantation hemodialysis strategy on early renal graft function was evaluated in 44 patients receiving hemodialysis in the 24 h preceding kidney transplantation and in 13 patients receiving hemodialysis more than 24 h before transplantation. The patients dialyzed less than 24 h before transplantation were stratified according to treatment with or without complement-activating dialyzers (cuprophane, bioincompatible membrane [BICM] versus polysulfone, biocompatible membrane [BCM]) and with or without ultrafiltration (UF). Serum creatinine (Scr) at days 0, 2, 5, 10, and 30, the time for Scr to decrease 50% (T1/2Scr), the incidence of acute renal failure (ARF; defined as urinary volume < 500 ml/d and/or necessity for posttransplantation hemodialysis), and early graft dysfunction (defined as T1/2Scr > 3.5 d) were registered. Scr was higher in BCM- versus BICM-treated patients (P < 0.0001 by variance analysis) and in patients receiving UF versus those receiving no UF (P = 0.0009). T1/2Scr was higher in treatment with BICM versus BCM (7.4 +/- 7.9 versus 3.1 +/- 2.9 d; P < 0.05) and UF versus no UF (7.1 +/- 7.7 versus 2.7 +/- 2.0 d; P < 0.01). The evolution of Scr was markedly more favorable in the patient group treated with BCM without UF (T1/2Scr 1.7 +/- 0.8 d) compared with the group treated with BICM and UF (T1/2Scr 9.3 +/- 9.1 d; P < 0.01). The remaining groups (BICM without UF and BCM with UF) showed intermediate results. The incidence of ARF and early graft dysfunction was higher in the group on BICM with UF compared to BCM without UF. Functional differences persisted up to 1 mo after transplantation. Patients who underwent dialysis with UF more than 24 h before transplantation had a more beneficial evolution of renal function parameters than those who were dialyzed with UF less than 24 h before transplantation. In conclusion, the use of BICM and the application of UF within 24 h before kidney transplantation enhance the risk of posttransplantation ARF and early graft dysfunction.

Rifampicin-associated acute renal failure: pathophysiologic, immunologic, and clinical features.

A 71-year-old woman was treated for a relapsing pulmonary tuberculosis with reinstitution of rifampicin after a medication-free interval of 2 years. After ingestion of the second dose, she developed severe hemolytic anemia and acute renal failure (ARF) necessitating dialysis. We demonstrated the presence in the patient's serum of rifampicin-dependent immunoglobulin G (IgG) and IgM antibodies, which caused red blood cell lysis through interaction with the I antigen on the erythrocyte surface. A review of the literature yielded 48 cases of rifampicin-associated renal failure. A subgroup of 37 patients could be distinguished, which, analogous to our case, suddenly developed ARF and frequently also developed hemolytic anemia and/or thrombocytopenia during intermittent or interrupted treatment. Regarding the pathogenesis of the ARF, renal biopsy consistently revealed tubular lesions. Although intravascular hemolysis with hemoglobinuria may play a role, it is not uniformly present. Our demonstration of an antibody with anti-I specificity provides a possible explanation. The I antigen is also expressed on tubular epithelium and may, therefore, be the target structure through which rifampicin-antibody complexes lead to tubular cell destruction. The other cases of rifampicin-associated ARF were unrelated to this subgroup: two cases of rapidly progressive glomerulonephritis, five cases of acute interstitial nephritis, and four cases of light chain proteinuria were recorded.

Leukocyte CD14 and CD45 expression during hemodialysis: polysulfone versus cuprophane.

The expression of CD14 on monocytes and CD45 on monocytes and granulocytes was evaluated during hemodialysis by flow cytometric analysis in the 'resting state' and after in vitro stimulation with phorbol myristate acetate (PMA). A comparison of complement activating cuprophane (CU) versus less complement activating polysulfone (PS) was undertaken. 'Resting state' CD45 expression on granulocytes increased markedly during CU dialysis compared to time 0, whereas this rise was only moderate with PS (CU vs. PS, p < 0.01). When considering the increase in expression upon PMA stimulation, a lower value was obtained during CU dialysis for both CD14 (monocytes at 60 and 240 min) and for CD45 (monocytes and granulocytes at 15 min). In conclusion, granulocytes in the 'resting state' expressed more CD45 on their cell membranes during CU dialysis, whereas CD14 and CD45 upregulation after ex vivo addition of PMA was blunted during CU dialysis.

Recombinant hirudin: a specific thrombin inhibiting anticoagulant for hemodialysis.

The first experience with hirudin as an alternative anticoagulant for heparin in hemodialysis is reported. Recombinant hirudin (HBW 023) was administered in 20 patients as a bolus before dialysis with low flux polysulfone dialyzers (PS400), the dosage being adapted stepwise from patient to patient by 0.02 mg/kg to the occurrence of clotting or bleeding. Four different administration schedules were studied. The first three schedules (0.02 mg/kg, N = 1; 0.04 mg/kg, N = 1; 0.06 mg/kg, N = 4) were discontinued because of clotting. The 0.08 mg/kg schedule was maintained without clotting event in 14 patients. Bleeding was not observed. Plasma hirudin averaged 503.9 +/- 214.0 and 527.7 +/- 217.1 ng/ml after two and four hours of dialysis, and decreased during an interdialytic interval of 44 hours to 223.2 +/- 86.2 ng/ml. Modified antithrombin III (P < 0.05) and activated partial thromboplastin times were lower (P < 0.01) under hirudin compared to heparin; these coagulation parameters were closer to normal during hirudin treatment. The patients developing clotting could be distinguished from those without clotting by the registration of the activated clotting times (9.2 +/- 3.0 vs. 18.7 +/- 3.2 min after 2 hr, P < 0.01; 8.1 +/- 3.0 vs. 16.2 +/- 3.8 min after 4 hr of dialysis, P < 0.05); cut-off value below which clotting is to be expected was 12 min). It is concluded that administration of hirudin as a bolus before the start of dialysis, at a dosage of 0.08 mg/kg, is not complicated by clotting or by bleeding. Coagulation tendency can optimally be monitored by the registration of the activated clotting time.(ABSTRACT TRUNCATED AT 250 WORDS)

Assessment of urea and other uremic markers for quantification of dialysis efficacy.

To validate azotemic markers as an index for intradialytic changes in solute concentration, we compared eight solutes (pseudouridine, xanthine, hypoxanthine, peak 4, peak 5, p-hydroxyhippuric acid, indoxyl sulfate, and hippuric acid) with five classical azotemic markers (urea, creatinine, uric acid, phosphate, and potassium). We determined concentrations by reversed-phase HPLC coupled to ultraviolet absorption or photometrically. Seven compounds showed significant intercorrelation (P less than 10(-5)): urea, pseudouridine, uric acid, peaks 4 and 5, p-hydroxyhippuric acid, and creatinine. The hippuric acid concentration change after dialysis correlated with the change for these seven compounds and also with indoxyl sulfate, hypoxanthine, potassium, and the group of unidentified ultraviolet-absorbing HPLC peaks accumulating in uremia. We conclude that urea only partially represents the concentration changes of other retention compounds after dialysis; alternative markers, e.g., hippurate, should be considered.

Benefit of intravenous essential amino-acids in catabolic patients on chronic maintenance hemodialysis.

The effect on the clinical status of catabolic hemodialysis patients of I.V. essential amino-acids administered over 3 months at the end of each dialysis is assessed in an open clinical follow-up study of 10 patients; these patients showed a progressive deterioration of general condition and a progressive weight loss in the period before the start of the treatment. The study was undertaken in a hospital dialysis unit, with as main outcome measures body weight, hematocrit, a scoring index of general condition and degree of edema. In patients showing a progressive and consistent loss of body weight in the months preceding the study, after the first treatment month, body weight started to rise, increasing after 3 months from 56.2 +/- 2.3 to 58.6 +/- 2.4 kg (p less than 0.01). The hematocrit raised from 22.4 +/- 1.6% up to 26.5 +/- 1.9% (p less than 0.02). Over this period, only 2 liters of packed cells were administered, in contrast to an overall need of 13 liters in the preceding 6 months. Peripheral and/or pulmonary edema disappeared. A scoring index, of general condition, increased from 5.1 +/- 1.5 before the start of the study to 11.7 +/- 0.8 after 3 months (p less than 0.01). It is concluded that the parenteral administration of amino-acids in catabolic patients on chronic hemodialysis has a beneficial effect on general condition, and the balance of body fluid and body mass.

Intradialytic body weight changes and dialyzer pore size as main contributing factors to the evolution of beta-2-microglobulin in dialysis.

Cuprophane hemodialysis is associated with an early fall of leukocyte counts and an intradialytic rise in serum beta 2-microglobulin (beta 2M), in contrast to dialysis with more compatible dialyzers. It has been suggested that these two phenomena may be related. This study sets out to verify this hypothesis by comparing the evolution of leukocyte counts with that of beta 2M: (1) during dialysis with 5 dialyzer types with different pore size and/or leukocyte biocompatibility; (2) during first use and reuse of 3 dialyzer types, and (3) during sequential ultrafiltration and dialysis with cuprophane. In first-use dialyses, no relation could be found between changes in leukocyte counts and the evolution of beta 2M levels. Reuse of cuprophane and saponified cellulose ester resulted in a marked attenuation of the intradialytic fall in leukocyte counts after 15 min (change in white blood cell count: -72 and -17% for first-use and third-reuse cuprophane, -72 and -23% for saponified cellulose, respectively), but had no influence on the increase in beta 2M. Correlation studies of these data revealed that the intradialytic evolution of beta 2M was related to membrane pore size and, for membranes with a small pore size, to the intradialytic fluid losses: first-use cuprophane (p less than 0.05), saponified cellulose ester (p less than 0.001) and hemophane (p less than 0.01), and pooled first-use and reuse cuprophane and saponified cellulose ester (p less than 0.001). Cuprophane dialysis without ultrafiltration (dialysate Na+: 138 and 132 mEq/l) caused a fall in leukocytes, but induced no rise in beta 2M. Ultrafiltration with cuprophane either preceding or following dialysis consistently caused a rise in serum beta 2M, although a fall in leukocyte counts only occurred in the first case. Our data point away from a relationship between membrane biocompatibility, expressed as changes in leukocyte counts, and beta 2M concentration during hemodialysis. The major contributing factors appear to be dialytic fluid losses and membrane pore size.

Loss of residual renal function in patients on regular haemodialysis.

The literature offers scant data on loss of residual renal function in chronic haemodialysis patients. The present study was undertaken in 34 patients, to evaluate residual creatinine clearances (CCr) before the start of haemodialysis and after 3, 12 and 24 months. CCr progressively declined from 6.15 +/- 2.61 (before) to 1.40 +/- 1.29 ml.min-1 (after 24 months: p less than 0.01). The decrease was largest during the first three months of dialysis therapy (slope -0.99 +/- 1.01 ml.min-1.month-1 for the first three months vs. -0.23 +/- 0.12 ml.min-1.month-1 for the entire 24-month period: p less than 0.01). The decline in CCr during the first three months was significantly more pronounced in glomerular disease than in tubulo-interstitial disease (p less than 0.05). This could not be attributed to differences in blood pressure, body weight or hypotensive medications. Age and sex also had no influence. Our data indicate that there is a characteristic progressive loss of renal function in haemodialyzed patients and that the early decline is most pronounced in patients with glomerular disease. Regular assessment of residual renal function at least every three months is indicated in patients starting chronic haemodialysis treatment.

Challenge of phagocyte metabolism by extracorporeal circulation and membrane contact. A biocompatibility test.

In the present study, direct data on phagocyte metabolism during hemodialysis are collected by the determination of 14CO2-production from labelled glucose on whole blood samples. Fifteen minutes after the start of cuprophan dialysis, 14CO2 production per 10(3) phagocytes increased more than four-fold (from 77.6 +/- 30.2 to 315.0 +/- 69.6 DPM/10(3) phagocytes; P less than 0.01). Such a change was absent during dialysis with membranes containing polysulphone, AN69S or reused cuprophan. In vitro contact of normal blood with cuprophan membranes revealed an increase of 14CO2-production by a maximum of 9.0 +/- 3.4 X 10(3) D.P.M. at 15 min. (P less than 0.05). Such an increase was absent after contact with polysulphone. The measurement of 14CO2 production during glucose metabolism, by phagocytes present in micro-amounts of whole blood, is a valuable test for membrane biocompatibility. Cuprophan dialysis induces a challenge of phagocytic activity that is absent for dialysers containing other membranes.

Cardiac output-changes during hemodialysis with ultrafiltration.

Cardiovascular hemodynamics were studied noninvasively before, during and after hemodialysis with ultrafiltration in 18 patients on chronic hemodialysis. The cardiac output (CO) was determined by a continuous wave Doppler method. Overall, no major CO changes were seen (7.8 +/- 0.6 l/min post- versus 7.4 +/- 0.5 l/min pre-dialysis). Mean blood pressure rose slightly but significantly from 103 +/- 4 mmHg before to 113 +/- 3 mmHg after hemodialysis (p less than 0.01). Important interindividual differences in the intradialytic evolution of CO were observed. In patients with previous myocardial infarction or dilated cardiomyopathy (n = 12), CO rose significantly from 7.3 +/- 0.7 l/min before to 8.4 +/- 0.6 l/min after hemodialysis (p less than 0.05), while in patients without manifest myocardial disease (n = 6) CO decreased from 7.5 +/- 0.7 l/min to 6.6 +/- 0.9 l/min (NS). Comparison of the evolution of CO in both groups by variance analysis revealed a significant difference (p less than 0.01). It is concluded that, in response to hemodialysis with ultrafiltration, CO probably will increase in patients with myocardial infarction or congestive cardiomyopathy, but probably will decrease in patients without.