At Tokyo Women’s Medical University, Tokyo, Japan, we perform about 30,000 dialysis or plasmapheresis treatments every year. Characteristic features of blood purification therapy at our center include treatment of maintenance hemodialysis patients, whereas our center is one of the central medical organization in Tokyo, and frequent plasmaexchange and double-filtration plasma exchange, because 150～200 renal transplantats are performed at our center every year. Also, approximately 1% of blood purification therapy is performed in the Intensive Care Unit or in a regular room of the hospital ward. While we try our best to treat all our patients without incident or complication, blood purification therapy is required for a number of diseases, and these patients have a range of physical conditions. Variety of medical staffs participate this therapy, resulting in increasing possibility of incidents or accidents. To analyze reports from the incidents or the accidents, blood purification therapy may be performed safer than today.

Patients receiving continuous renal replacement therapy (CRRT) suffer not only from acute kidney injury (AKI), but also severe sepsis and/or multiple organ failure. Anticoagulation and nutritional management are important to improve the prognosis of these patients. In order to delay the clotting of the filter and CRRT circuit, anticoagulants are frequently used. However, circuit anticoagulation increases risk of bleeding. Therefore, the risks and benefits of intense anticoagulation must be considered. There is no direct evidence for optimal nutritional management in patients undergoing CRRT；therefore, nutritional management based on the management of AKI or sepsis is recommended in patients receiving CRRT. Sufficient administration of drugs, nutrition and blood transfusion is possible, because removal of unwanted solutes can be achieved through CRRT. On the other hand, a part of the administered nutrients such as vitamins and trace elements, as well as drugs are also removed. In CRRT, underfeeding may be permitted after assessment of the energy expenditure and nitrogen balance. Whenever possible, even if in little quantities, the patients should be fed enterally.

We outlined the adjustment of replacement fluid composition for continuous hemodiafiltration (CHDF) for the treatment of electrolyte disturbances. Commercially available replacement fluid includes 140mEq/L of sodium, 2mEq/L of potassium and 2.5mEq/L of ionized calcium. In patients with hypernatremia, adjusted relatively low sodium concentration (lower than the patient’s serum level and higher than 140mEq/L of serum) solutions should be used to avoid rapid correction-induced cerebral edema. Calcium-free replacement solutions should be used to correct hypercalcemia by CHDF, because commercially available replacement solutions include high levels of ionized calcium. In patients with hyperkalemia and hemodynamic instability, treatment with CHDF should be performed using potassium-free replacement solutions. In general, less than middle molecular substances are non-selectively removed by CHDF. Therefore, serum concentrations of electrolytes such as inorganic phosphate and magnesium should be carefully monitored in patients receiving long-duration and/or high-volume replacement CHDF.

Sivelestat sodium hydrate (SSH), a neutrophil elastase inhibitor, is used for the treatment of acute respiratory distress syndrome. However, acute renal failure may develop as a complication, and thus continuous renal replacement therapy (CRRT) is performed for many cases, leading to concerns about loss of the drug. Therefore, we examined the membrane permeability of SSH using a simulation circuit of CRRT. In the study, we added 20 mg of SSH to bovine blood or 4.5 L of saline, and circulated the solution in the CRRT simulation circuit for 2 hours, 3 times. Sampling was performed from 3 points, the blood return and removal sites and the filtrate, in 30-minute intervals. SSH concentrations were measured using HPLC. While SSH concentrations in the bovine blood on the side of filtration circuit were low regardless of the draining, SSH concentrations in saline increased on the sides of blood return and filtration circuit along with draining. SSH is highly protein bound, and thus does not permeate the membrane when it is in blood, even after 6 hours of CRRT. Therefore, no loss of SSH would be occur with CRRT, and does not pose a clinical problem.

【Objective】 Continuous hemodiafiltration (CHDF) using a hemofilter made from a membrane with cytokine adsorption properties is believed to be effective in removing cytokines during sepsis treatment. We devised a series double connection continuous hemodiafiltration (SD-CHDF) method for the treatment of hypercytokinemia and evaluated the clinical efficacy of this method. The method involves the serial connection of 2 polymethyl methacrylate membrane (PMMA) hemofilters, which increase adsorption area and result in increased cytokine removal capacity. 【Patients】 We examined 9 patients with increased serum IL-6 levels (＞900 pg/mL) who underwent SD-CHDF for hypercytokinemia at the intensive care unit of Almeida memorial Hospital from June 2010 to May 2012. We compared the clinical efficacy of SD-CHDF in our patients with that of single PMMA-CHDF (S-CHDF) in patients with a similar pathological condition. 【Methods and Results】 Serum IL-6 levels were measured before and 6, 12, and 24 h after the initiation of CHDF in both groups. Serum IL-6 levels decreased significantly at 6 and 12 h in the SD-CHDF group, whereas they decreased only at 24 h in the S-CHDF group. In addition, SD-CHDF significantly decreased blood lactate levels, whereas S-CHDF did not. 【Conclusion】 SD-CHDF using 2 PMMA hemofilters with cytokine adsorbing capacity is effective in the removal of cytokines during the treatment of hypercytokinemia. However, further investigation is necessary to confirm its efficacy in improving survival rates of septic patients.

The dosage of continuous renal replacement therapy (CRRT) is usually determined under the assumption that CRRT will be performed for 24 hr per day. However, CRRT frequently needs to be halted for several reasons in clinical settings. This study is aimed at evaluation of the actual treatment time of CRRT per day and the reasons for stopping CRRT in ICU. Seventy-two patients who needed CRRT in the ICU of Tokyo University Hospital were analyzed retrospectively. Each patient was treated using CRRT for 142.5±108.8 hr, whereas CRRT was halted for 4.72 ±3.05 hr. Reasons for stopping were the following: routine filter change (83%), medical procedure (7%), change access (3%), surgical procedure (1%), and miscellaneous other reasons (5%). Of note, routine filter changes required a shorter time (21.5±23.4 min) than other reasons. Although CRRT was held for less than 5% of the total treatment time in this study, recognizing the actual treatment time will enable determination of the optimal CRRT dosage prescription. Further studies must be undertaken to determine whether the CRRT stopping time has a meaningful impact on CRRT efficacy.

The aim of this retrospective study was to compare the lifetimes of two hemofilters---polysulfone and cellulose triacetate---to determine when sudden hypercoagulation occurred in continuous renal replacement therapy (CRRT). One group was set for no changes of the polysulfone hemofilter (PS) (number (n)=59）and the other was to change the cellulose triacetate hemofilter (CTA) (n=33) when the polysulfone hemofilter was coagulated. There were no differences in the basic characteristics of the patients who used PS and CTA filters. Also no differences were noted in hemofilter life-time and total doses of nafamostat before hemofilters were exchanged. However, the substitution flow rate (531±243mL/hr in PS group and 391±234mL/hr in CTA group) were different (P＜0.05). After exchange, 61.1％ of CTA filters were coagulation free during 24 hrs. In contrast, only 25.0％ of PS filters were coagulation free. Hemofilter life times were 18.6±14.5 hr and 27.6±13.1 hr in the PS and CTA groups, respectively. (P＜0.05). Judging from these results, the exchange of PS to CTA filter would be effective in case of coagulation.

Continuous Renal Replacement Therapy (CRRT) has become indispensable in the field of intensive care and hemofilters used in CRRT are required to be usable over a long period of time. In this study, hemofilter lifetime was examined in a crossover comparative study between AEF 0.7（Excelflo^®, membrane surface area 0.7m²）and AEF 1.0（Excelflo^®, membrane surface area 1.0m2）. In addition, the internal flow of the hemofilters was observed under fluoroscopic X-ray in an experimental setup. Lifetime with AEF 0.7 was 1,908±210 minutes-significantly longer than lifetime with AEF 1.0 (1,305±213 minutes). In addition, the experimental results suggested that the larger the membrane surface area, the more uneven the internal hemofilter blood flow distribution. AEF is designed so that the larger the membrane surface area, the smaller the L/D and the larger the header diameter. In this study, the decrease of linear velocity as a result of the increased header diameter and the effects of uneven blood flow distribution are thought to be the reasons why lifetime was shortened with AEF with a large membrane surface area. Accordingly, a proper L/D is important for extending hemofilter lifetime.

Blood purification therapy has become a widely used modality in critical care and intensive care. However, few studies have been carried out to investigate the influence of operating conditions and/or structure of the hemofiltration membrane on the solute removal efficiency. We examined the effect of the total flow rate of dialysate flow and filtration flow (Q_D+Q_F), the ratio of dialysate flow rate to filtration flow rate (Q_D/Q_F) and the hemofilter pore size on clearance, using the pore model and mass transfer model for hemodiafiltration. Clearance was increased with increasing Q_D+Q_F and pore size. Clearance of small molecular weight solutes showed a minimum value by changing Q_D/Q_F, whereas for middle or large molecular weight solutes a maximum value was shown. Furthermore, clearance showed a minimum or maximum value at different Q_D/Q_F depending on the solutes.

In this study, we report the case of blood leakage into a coupler (supply site of dialysis fluid) during an extracorporeal ultrafiltration method (ECUM). An alarm did not notify blood leakage even after 65 min since the start of treatment. To identify ways of detecting blood leakage rapidly, we examined leakage at the inlet site, intermediate part, or exit site. As a result, we found a considerable delay in the response time of the alarm notifying a blood leakage, which may occur because of several factors such as site of leakage, condition of the coupler, and direction of blood flow because blood leakage is influenced by sedimentation rate of gravity and temporary pushing of a circuit block through the connector using saline. During ECUM, for continuous monitoring and rapid detection of the blood leakage, the coupler should be placed in a position such that for an oncoming direction of dialysis fluid, the blood flows in an anti-gravity direction (bottom to top), as in the case of hemodialysis, and in the case of a parallel direction of the dialysis fluid, the blood flows in the direction of gravity (top to bottom). Furthermore, for safety, a sensor with self-test capability will be retrofitted in dialysis machines, thereby enabling detection of blood leakage in any part of the dialyzer.

Recently, off-pump coronary artery bypass grafting (OPCABG) without cardiopulmonary bypass has become less stressful surgery for coronary artery bypass grafting (CABG). There have been a lot of reports discussing risk factors involving in an association between on-pump coronary artery bypass grafting and acute kidney injury (AKI) requiring continuous renal replacement therapy (CRRT), however, there are few papers on OPCABG. The purpose of the present study was to assess the impact of OPCABG on the incidence of AKI requiring CRRT. An observational study of 237 consecutive non dialysis patients who underwent isolated CABG using OPCABG was conducted. Among them, 33 patients needed CRRT due to AKI. Variables with a P＜0.05 in bivariate analysis collected from pre-, intra- and postoperative data were entered in the multivariate and proportional hazards regression analysis as independent AKI requiring CRRT risk factors after OPCABG were carried out. The risk factors that were independently associated with AKI requiring CRRT were：pre-estimated glomerular filtration rate (GFR) (less than 60mL/min/1.73m²), pre-serum albumin level (less than 3.5g/dL), pre-hemoglobin level (less than 12g/dL), intra-urine volume(less than 600mL/hr), use of intra-aortic balloon pump, and post-PaO₂/FiO₂ (less than 300). In conclusion, the risk of developing AKI requiring CRRT depended on the levels of GFR, serum albumin and hemoglobin before surgery, on the levels of urine volume and use of intra-aortic balloon pump during surgery and the levels of PaO₂/FiO₂ after surgery.

In hemodialysis therapy, prescribed blood flow rate (Qb, mL/min) is an important factor that determines the amount of waste removal. It has been reported, however, that actual blood flow rate (aQb, mL/min) is influenced by the needles used for hemodialysis and that aQb is usually lower than Qb. In the present study, we compared five types of structurally different double lumen catheters (DLCs) to evaluate which types of DLC minimize the difference between Qb and aQb. Because the arterial and venous sides of DLCs differ in shape, both sides were evaluated separately. At the arterial access site, Flexxicon^®Ⅱ showed a largest decrease (80.4±0.01mL/min), whereas Gam Cath^® showed a minimum decrease (23.8±0.55mL/min). At the venous access site, all DLCs showed a reduction by less than 6.58±2.29mL/min. These results indicate that the structure of DLCs influences aQb and the effect is more clearly shown on the arterial side. Difference of aQb has the great influence by the structure of the arterial site, or the difference of arterial cross sectional areas. Especially when performing a setup of a high flow, it is necessary to determine Qb in consideration of the feature of various DLC.

We did not have a system of rapid response to problems in continuous blood purification therapy (CBP) in children when we started to perform this therapy in the intensive care unit (ICU) of our hospital. Therefore, we retrospectively analyzed the results of CBP in children in the ICU of our hospital. We performed CBP for a total of 104 days in seven children over a three-year period. For patients whose body weight was less than 20kg, red cell concentrates and fresh frozen plasma were used as the priming solution. Diafiltration was performed before the start of extracorporeal circulation to correct electrolytes and to warm the circuit. Exchange of the circuit was performed every 24hours by connecting the current circuit to the new circuit in series. Seventy-four adverse events occurred in total. The most frequent adverse event was intra-circuit coagulation (34events). However, we have learned many troubleshooting techniques with the accumulation of experience. We can now smoothly cope with such problems because we created a manual and we have workshops where we discuss any problems that occurred. The anxiety of our clinical engineers has decreased. Stable blood purification therapy cannot be achieved with intra-circuit coagulation. Therefore, it is necessary to deal with intra-circuit coagulation immediately.

We conducted a retrospective study to determine the impact of acute blood purification strategies on patient outcome. Sixty-one critically ill patients who required acute blood purification between January and December 2011 in our institution were included. Data on patient characteristics and disease severity were obtained through medical records and the prognostic factors determined. Severity was evaluated using APACHEⅡscores, the number of organ dysfunctions, the presence of sepsis, and the affected organ. The severity of acute kidney injury (AKI) was evaluated using the RIFLE classification. Patients were divided into a survival or non-survival group, according to whether they survived for more than or less than 90 days, respectively. In the non-survival group, the morbidity of multiple organ dysfunction syndrome and the number of failed organs were significantly higher than in the survival group. Furthermore, in subanalysis of patients with AKI, the continuous renal replacement therapy (CRRT) group, which was initiated with CRRT, had higher APACHEⅡscores and a lower survival rate compared with the intermittent renal replacement therapy (IRRT) group, which was initiated with IRRT. In the multivariate analysis of prognostic factors, the APACHEⅡscore was identified as a meaningful independent factor. Taken together, the findings suggest that the APACHEⅡscore can be used as an independent prognostic factor of patient survival, even in patients who require acute blood purification therapy.

Objective：To assess the effect of endotoxin adsorption therapy on patients with acute exacerbation of refractory drug-resistant interstitial pneumonia. Methods：Subjects were patients diagnosed with interstitial pneumonia whose respiratory failure was not ameliorated after steroid pulse therapy. Subjects were divided into two groups, treated or untreated with PMX-DHP, and statistical comparisons were performed between the two groups. Results：In the PMX group (n=11), compared with the non-PMX group (n=13), oxygenation (pre PMX vs. post PMX, P/F ratio：120.6＋/－43.5 vs. 183.6＋/－87.1, p＝0.044) and Week 4 prognosis (Kaplan-Meier method, logrank test, p=0.037) improved significantly. Conclusion：PMX-DHP can reverse serious acute respiratory failure in patients with acute exacerbation of drug-resistant interstitial pneumonia.

We present a case of acute liver failure and platelet transfusion refractoriness that was successfully treated with plasmapheresis. A 43-year-old female patient developed pancytopenia after chemotherapy for acute myeloid leukemia. She was admitted to our ICU because of sepsis on 30 day after the completion of chemotherapy. On the 3rd ICU day, the patient developed acute liver failure that was diagnosed by findings of marked elevation of transaminase and serum bilirubin and prolongation of international normalization ratio of prothrombin time. Three times of plasmapheresis were immediately performed and this therapy resulted in rapid recovery from derangement of liver function. Since the patient received frequent platelet transfusion due to bone marrow suppression, platelet transfusion refractoriness occurred and then thrombocytopenia-related pulmonary alveolar hemorrhage occurred on the 10th ICU day. To restore the efficacy of platelet transfusion, three times of plasmapheresis were performed during a period of 48 hours for elimination of platelet-related alloantibodies. An increase in platelet count by platelet transfusion was restored after consecutive plasmapheresis and resulted in improvement of pulmonary alveolar hemorrhage.Plasmapheresis therapy is a useful option for treatment of diseases for which the pathogenesis is possibly associated with relatively high molecular substances such as allo- or auto- antibodies.

When CBP (continuous blood purification) is performed with PCPS (percutaneous cardiopulmonary support), blood can be drawn from the venous line of PCPS. However, there might be a risk of injecting bubbles. We examined the risk of injecting bubbles into a patient when a circuit of CBP is connected to the venous line of PCPS. We found that when the circuit is completely transected or there is a loose connection, the oxygenator cannot trap air bubbles, and the bubbles were rapidly injected into the patient. To avoid this risk, the vascular access of PCPS and CBP should therefore be separate.