Continuous veno-venous hemofiltration for ARF in critically Ill patients
The mortality of critically ill patients who develop ARF in an ICU setting is extremely high (50-80%). Any mode of renal replacement therapy chosen should be able to achieve solute and water clearance while maintaining hemodynamic stability, have a positive effect on nutrition, and have low complication rates. AIM: To determine the efficacy and feasibility of Continuous Venovenous hemofiltration (CVVH) in critically ill patients with ARF. Inclusion criteria: Patients with ARF requiring 2 or more inotropes to maintain systolic blood pressure >100 mm of Hg. Failed or technically impossible hemodialysis or peritoneal dialysis. Time Period: July 2002 - June 2003. MATERIALS AND METHODS: Polysulfone hemofilter 0.7m2, [Aquamax (Edwards) or Multimat BL680 (Bellco).] Blood flow of 150-200 ml/minute (Travenol). Volumetrically controlled Ultrafiltration of > 2000 ml per hour (Watson Marlowe) and replacement fluid infusion [(Infusomat-P) post filter]. Anticoagulation: Heparin infusion or regional heparinisation. RESULTS: 22 patients included, 6 with recent abdominal surgery. 11 underwent hemodiafiltration and hemofiltration each. Severe sepsis was present in 21, and DIC in18. 5 patients were on immunosuppressive therapy. The time from ICU admission to the start of CVVH was 114 + 88.08 hours. The duration of CVVH was 35.93 + 20.91 hours, (range 11 to 84 hours). The mean hourly ultrafiltration of 93.72 + 65.57 ml and total ultrafiltration of 3955.55 + 4132 ml was tolerated by all patients without limiting hypotension. The APACHE II scores had significantly worsened between admission (22.5 + 6.71) to starting CVVH (36.05 + 4.08), [P<0.001]. The daily costs of CVVH were Rs. 5000 compared to Rs. 2150 for PD and Rs. 1500 for extended daily dialysis CONCLUSIONS: CVVH was effective in providing metabolic correction in ARF, in the setting of multi-organ failure. It is technically feasible even when conventional hemodialysis or peritoneal dialysis cannot be performed.
Keywords: ARF, venovenous hemofiltration, freestanding, ultrafiltration, critical
Acute renal failure in seriously ill patients in the intensive care unit has been shown to be an independent predictor of morbidity and mortality. Most of these patients have multiorgan failure, severe sepsis, DIC, a hypercatabolic state, and a requirement for very large fluid volumes to maintain their nutritional status and correct coagulation abnormalities. The problem is worsened by the need for higher doses of inotropes to maintain a stable circulation and the fact that most of them are anuric, with a propensity to develop pulmonary edema.
Intermittent hemodialysis may be impossible in such patients, because of their labile circulation, and they may require a slow continuous therapy to remove enough fluid and solute load daily while maintaining hemodynamic stability. Several techniques of continuous renal replacement are available with dedicated machines built specifically for this purpose. Most of these machines are very costly, and are closed systems utilizing disposables manufactured only by the same company. Free standing components from old dialysis machines can be used to perform continuous venovenous hemofiltration or hemodiafiltration, a technique known as adaptive technology. In our unit a combination of free standing pumps and monitoring devices modified to accommodate various disposable hemofiltration kits are in use and we describe the use of this technology in performing continuous renal replacement therapies.
To determine the efficacy and safety of continuous veno venous hemofiltration for acute renal failure in critically ill patients.
22 patients were studied between July 2002 and January 2004.
Patients with oligo anuric ARF having Systolic Blood pressure = 100 mm of Hg with > 2 inotrope drugs 3 or more organ failures. Failed or technically impossible peritoneal or hemodialysis.
Need for renal replacement therapy according to conventional indications.
Patients successfully treated by conventional intermittent hemodialysis.
Organ failure was defined according to criteria described by Mehta (2001)
Illness Severity was scored according to APACHE II scoring system at admission and initiation of CVVH.
Blood urea (BUL), and Serum creatinine were estimated daily. Electrolytes, blood gases and aPTT were estimated every 4 hours during the procedure and the composition of replacement fluid and dose of anticoagulant modified accordingly.
Results were analysed by students paired 't' test
Polysulfone hemofilters having a surface area volume of 0.7 m2 and an ultrafiltration coefficient of 33 ml/mm of Hg/hour [Aquamax (Edwards life sciences) and Multimat BL 680 (Bellco)] were used to perform hemofiltration.
Temporary access was secured with 12F double lumened cannulae, using the femoral route in 14 patients and the internal jugular in 8. Blood flows of 150 - 175 ml/minute were obtained using a Travenol blood pump calibrated with saline before each procedure and daily during each procedure, flow variations were not > 5%. Volumetrically controlled ultrafiltration of > 2000 ml/hour was obtained using a refurbished Watson Marlowe pump, calibrated as described above using Aquamax tubing sets.
Dialysate flows of 1000 ml/hour, countercurrent to blood flow, were obtained with peritoneal dialysis solution and peristaltic pumps (Infusomat B. Braun) and externally warmed replacement fluid was infused, pre and post dilution, at 1700 - 2000 ml per hour using peristaltic pumps (Infusomat B. Braun). Replacement fluid consisted of normal saline and lactated Ringers solution with added calcium gluconate, sodium bicarbonate and dextrose. Continuous heparin infusion at 250 units/hour or regional heparinisation was carried out using syringe pumps. (B. Braun)
22 patients underwent 24 sessions of continuous renal replacement therapy between July 2002 and January 2004, 11 receiving hemofiltration alone and 11 hemodiafiltration. 1 patient received 1 session each of hemofiltration and hemodiafiltration. The profile of patients and indications for starting hemofiltration are shown in [Table - 1]. As can be seen in [Table - 2], there was a significant worsening in the APACHE II scores from admission in the ICU to initiating CVVH. This could be attributed to the very severe illness that this patient group had and the attempts to manage them with more conventional treatments initially. We noted a significant improvement in blood urea, serum creatinine, acidosis and hypocalcemia in these patients, while sodium and potassium were not significantly changed, [Table - 3] however all our replacement fluids contained some potassium. Hypotension was the commonest adverse event in our patients, but the overall incidence was less than 1 per treatment, while no patient had line disconnection or air embolism. [Table - 4] Circuit clotting contributed to stopping the treatment in 13 patients, while 3 had the treatment discontinued in accordance with the relatives or treating physicians decision. 5 patients subsequently were switched over to daily and then alternate day hemodialysis until their discharge from the hospital or recovery from ARF. 17 patients died in hospital without recovering from ARF, 3 of whom had discontinued treatment, giving an overall in hospital mortality of 77.27%. [Table - 5] This was not significantly different from the predicted mortality of 83.8% based on the APACHE II scores and the regression equation (P = 0.1).The 30 day mortality calculated from the day of diagnosis of ARF was 77.27%, with 4 patients making a complete recovery from ARF, while 1 patient who recovered from ARF died 43 days after hospital discharge at home, probably due to underlying Carcinoma of the colon. The treatment was associated with failure to achieve ultrafiltration despite maximum pressor support in 2 patients while 1 had worsening hyperkalemia during the procedure leading to it being abandoned. The daily material costs of the procedure are as shown in [Table - 6], compared to Rs. 2150 for peritoneal dialysis and Rs. 1500 for a 12 hour hemodialysis session. The cost of the procedure would increase by around Rs. 30 for every additional 1 liter increase in exchange volume. All materials are supplied to patients after being purchased at special hospital rates and the institution does not charge for the procedure as no capital costs have been incurred in building our own apparatus.
In our study, the patients entering continuous renal replacement therapy were the patients with the worst severity of illness scores. This was inevitable because of the study design which excluded patients who could be taken for more robust forms of renal replacement therapy. In 12 patients, previous attempts at conventional intermittent hemodialysis or peritoneal dialysis had proved inadequate or unsuccessful, while 5 patients had had recent abdominal surgery, making peritoneal dialysis impossible. Higher SAPS-II scores and organ dysfunction in patients selected for CRRT as compared to intermittent therapy have been noted by other workers in a prospective randomized multicentre study. Continuous veno venous hemofiltration (CVVH) has been shown to provide better control of acidosis and uremia than peritoneal dialysis in acute renal failure due to malaria and sepsis in a study from Vietnam, improve survival, and reduce the risks of dying. We also noted a very significant reduction in blood urea and creatinine, and significant improvement in serum calcium and bicarbonate, providing excellent metabolic control. Although serum potassium and sodium did not significantly change over the entire study range, the patients in whom the treatment was initiated for hypo or hypernatremia or hyperkalemia complete correction was achieved in all but 1. The maintenance of sodium and potassium within a very tight range [Figure - 1] and [Figure - 2], reflects the intensive monitoring of electrolytes every 4 hours and the tailoring of replacement fluid accordingly. The better tolerance of slow ultrafiltration is borne out by the fact that 2 of our patients had were able to stop inotropic support during the procedure, 7 required lower doses and no of drugs, 8 had no change in inotrope requirement despite ultrafiltration, and only in 5 patients did the requirement for inotropes increase markedly or lead to ultrafiltration failure. In contrast only 1 patient achieved sufficient ultrafiltration to accommodate her requirement for blood products, drugs and nutritional support and still have an improved arterial oxygenation on peritoneal dialysis, while hemodialysis had to be stopped in 6 patients because of inability to achieve adequate ultrafiltration without severe hypotension. A temporary stoppage of ultrafiltration or decrease in its rate was found to improve blood pressure in those patients who had a transient fall, and only one procedure had to be terminated because of hypotension. That lower rates of ultrafiltration and greater frequency are tolerated better, and are associated with improved renal recovery has also been shown previously and the one unequivocally proven benefit of CVVH has been the ability to maintain better hemodynamic stability with larger fluid removal. The average filter life of 35.93 + 20.91 hours [Table - 7] is slightly lower than expected and was the main reason for stopping the procedure as the high cost of filters does not allow the daily changes recommended by other workers. A fear of life threatening bleeding led to a use of lower doses of heparin (average 9500 Units/day) than in most other studies (21000 Units/day) and may have contributed to filter loss. In this study, we used freestanding pumps of old dialysis machines modified to function with dedicated tubing sets for hemofiltration, a technique described as adaptive technology. Safety monitoring was ensured by use of alarm systems present on the machines and also very low threshold settings on the ICU multipara monitors. The incidence of serious complications was low and we noted no life threatening complications. In contrast 1 patient had a cardiorespiratory arrest on conventional hemodialysis, despite the use of a modern volumetric machine. This indicates the safety of the procedure in very hemodynamically unstable patients who are unsuitable for other forms of renal replacement therapy. That CRRT patients constitute the patients with the poorest prognosis has been reported before, even in randomized controlled trials.
The in hospital mortality of our patients was slightly lower than that predicted by logistic regression analyses based on the APAPCHE II scores and the disease coefficients of Knaus and Draper, although the reduction did not reach statistical significance. The value of 77.27% is similair to the 79% reported by Guerin. In the recently published multicentre Programme for Improvement of care in acute renal disease (PICARD) study, the overall mortality of patients selected for CRRT was 70%, reaching 100% in 2 of the centers and 90% in 1 centre for patients who were switched over to CRRT after failing intermittent therapies. The high mortality of patients taken for CRRT reflects the fact that they frequently constitute the most severely ill patients or those who worsen on other therapies, as has been noted earlier. The costs of CVVH are almost 3 times as high as for extended daily dialysis and two and a half times as high as for peritoneal dialysis, even neglecting recovery of the machines initial cost. If the filter and vascular access life are prolonged, the daily cost actually decreases, even though the actual procedure costs increase largely from the costs of replacement fluids. The optimum balance seems to be obtaining a filter life of 60 hours, as compared to the average life of 36 hours in our study, caused probably by use of lower doses of anticoagulant in our patients, a large number of whom had DIC. The maximum life of one of our filters was 84 hours, which did not lower the cost much further. If we had followed the recommendations of the Australian workers of daily filter changes the average daily costs would have increased to around Rs. 7300, which is too high for our patients. Manns et al have reported that CRRT results in an improved rate of renal recovery among survivors as compared to intermittent HD, and the lack of need for long term renal replacement therapy offsets the higher short term costs. It has been suggested that CRRT with its higher costs should be reserved for certain select patient populations, which was the aim of this study, where only patients intolerant of intermittent therapies were taken for CVVH.
Continuous venovenous hemofiltration can be performed using modified hemodialysis machine components and available hemofilters and tubing sets. It results in hemodynamically well tolerated ultrafiltration, good control of uremia, and acidosis, and is an effective treatment of pulmonary edema. The procedure has a low risk of complications and can be done in patients with multi system disease even where other forms of renal replacement therapy are not possible or have failed. Costs still remain a major limiting factor, hence the procedure is often resorted to late and results may be improved if it is used earlier in the course of the disease. Despite the hemodynamic instability and organ system failures, the patients who survived to hospital discharge had complete renal recovery and no requirement for long term renal replacement therapy.
The authors gratefully acknowledge the invaluable help received during the study from Dr. B. D. Bande, Dr. S. N. Kalashetti, Dr. Vandita Shah, Dr. A. V. Godbole, Dr. S. P. Jagtap, all ICU nursing staff and resident doctors, dialysis technicians, and Mr. M. P. Registrar, and Mr. R. Dhawale of the department of biomedical engineering.
[Figure - 1], [Figure - 2][Table - 1], [Table - 2], [Table - 3], [Table - 4], [Table - 5], [Table - 6], [Table - 7]