Indications, Contraindications, and Complications of Peritoneal Dialysis in Acute Renal Failure

Outcomes

Regarding patient outcomes, a number of prospective and retrospective studies and some randomized studies have compared mortality rates of patients with AKI when treated with PD versus with HD or CVVHD. Although the patient populations in the various studies and treated groups are not exactly the same, and there is some selection bias in patients that may favor PD, most studies have shown outcomes to be at least as good for PD as for HD, as shown in Table 179.1 .

Except for the study by Phu et al., all studies summarized in Table 179.1 have shown that in patients with AKI treated with PD, mortality and incidence of renal recovery are roughly equivalent to those in similar patients treated with HD. Firmat and Zucchini, reviewing literature reports involving more than 1100 patients, concluded that the mortality rate was identical for patients with AKI receiving PD and those receiving HD. Most of these studies were performed in the 1970s and 1980s and many used acute PD catheters rather than chronic tunneled catheters and open systems for draining the dialysate. However, Struijk continued analyzing patients treated with each modality at the Academic Free Hospital in Amsterdam from 1986 through 1999, using tunneled catheters and closed drainage systems (DG Struijk, personal communication, 2000). In his studies, mortality in patients treated with PD was identical to that of patients treated with HD. In the study by Phu et al., 70 adult patients with AKI, because of severe falciparum malaria in 48 and sepsis in 22, were assigned randomly to treatment by PD or by continuous venovenous hemofiltration (CVVH). The mortality was significantly higher in the group treated with PD (47%), as well as a lower rate of renal recovery than in the group treated with hemofiltration (mortality 15%). The PD schedule was very aggressive (70 L of fluid per day) and was performed using acute PD catheters and open drainage containers. Urea clearance of PD was about equal to that with CVVH, but creatinine clearance with PD was about half that with CVVH. What is most unusual about the Phu study is the exceedingly low mortality of the group treated with CVVH, rather than an unusually high mortality in the group treated with PD. As Daugirdas pointed out, it is possible that the heparin anticoagulation of CVVH was of benefit to the many patients with malaria in this study. Also, hyperglycemia accompanying PD may have stimulated malarial growth in the liver or red cells, or high osmolality may have diminished white cell function. Failure to correct acidosis, a serious problem in the PD group, may have been due to use of acetate as buffer in the PD solution, as opposed to lactate in the CVVH infusion fluid.

Gabriel performed a randomized study of patients with AKI, treating 60 patients with daily HD and 60 with PD. Patients were excluded if they had absolute contraindications to either therapy. Peritoneal access was a Tenckhoff catheter placed by trocar by a nephrologist. Mortality over a 30-day period was almost identical in the two groups. George randomized 50 patients with AKI, 25 to PD and 25 to CVVHD. Access was by a rigid, acute PD catheter. Mortality was statistically the same in both groups. Ponce randomized 143 patients with AKI to treatment by PD or extended daily HD. In-hospital mortality was identical in the two groups and in fact odds-ratio of mortality favored patients treated with PD.

Several of the studies summarized in Table 179.1 were randomized and prospectively controlled, and yet there was still some bias in patient selection. However, this bias worked for and against patients treated by PD. Patients with AKI who have abdominal trauma, are awaiting abdominal surgery, or have abdominal drains or severe ileus cannot be treated by PD. In general, however, patients with AKI after surgery have a higher rate of recovery from AKI than patients with other causes of AKI, such as sepsis and shock. In these studies, PD often was chosen for patients with hypotension or cardiovascular instability that would make HD dangerous, a practice that also selected a group with a potentially worse outcome. The techniques used for PD in most of the studies were antiquated by today's standards. In many studies including that by Phu, semirigid acute PD catheters were used rather than Tenckhoff catheters. Acute PD catheters have irregular outflow characteristics, and most are removed and re-inserted every 3 days. Each insertion increases the risk of bowel puncture and outflow failure. PD fluid was infused from bottles in many of the studies, and there were no Y-sets to allow drainage and infusion of PD fluid though a single catheter connection. In spite of the use of rather crude techniques in many of these studies, PD patients recovered renal function and survived at least as frequently as patients treated with HD, with the notable exception of the study by Phu et al.

Renal Recovery

In many of the studies comparing PD with HD for AKI, the improved survival in the PD group correlated with a higher rate of renal recovery. Of all patients with AKI few recover general health but fail to recover renal function, leaving the hospital with ESRD to be supported by dialysis (less than 10%). In patients with ESRD, treatment by continuous ambulatory PD (CAPD) results in better preservation of intrinsic renal function than treatment with HD. This preservation of renal function is important in ESRD because it maintains the endocrine function of the kidneys, diminishes the clearance requirements for adequate dialysis, minimizes required ultrafiltration during dialysis, and thus diminishes physiologic stress during dialysis. Intermittent HD is known to have the following nephrotoxic effects: (1) generation of inflammatory mediators by the extracorporeal circuit, (2) concomitant and rapid decrease in osmolality and vascular volume, which diminish renal perfusion, and (3) hypotensive episodes, which result in fresh ischemic lesions in the kidneys.

By contrast, the effects of CAPD therapy help maintain renal perfusion, because of the following: (1) smaller daily variation in body weight, (2) more constant blood pressure, (3) continued mild overhydration, with higher mean pulmonary arterial pressure, (4) persistent high blood osmolality, partly because of glucose, and (5) continued removal of proteins from the blood (including β ₂ -microglobulin, albumin with associated uremic toxins, plasminogen activator inhibitor type 1, and immunoglobulins).

Given the beneficial effects of PD, it is not surprising that some patients started on PD for what appears to be ESRD recover intrinsic renal function and no longer need dialysis (3.3%). Given the negative physiologic effects of HD, it is also not surprising that very few patients with ESRD who are treated with HD recover renal function (0.8%). Recovery of renal function is most common in patients whose renal failure was caused by uncontrolled hypertension, cardiac failure, nephrotic syndrome, rapidly progressive renal failure, analgesic nephropathy, urinary obstruction, or cholesterol emboli. Many of these underlying conditions are better corrected by CAPD than by HD, because of continuous chemical removal, better preservation of renal perfusion and glomerular filtration rate, and slow removal of immunoglobulins. These same physiologic and chemical benefits may account for the higher recovery of renal function, in most studies, in patients with AKI treated with PD than with HD.

Complications

There is general consensus that continuous dialysis therapies such as CVVH and CVVHD are the most efficient therapies for AKI, with the fewest adverse physiologic effects. These “gentle” forms of therapy remove fluid at a slow rate and cause less decrease in cardiac output than HD, especially when HD is performed for only 3 to 4 hours daily or every other day. CVVH and CVVHD do not affect pulmonary function adversely or significantly activate the complement cascade. Continuous arteriovenous hemofiltration (CAVH) was described first in 1967. Pump-assisted CVVH and CVVHD were developed to make the rate of blood flow through the hemofilter more consistent, improve clearances, eliminate problems and risks of arterial access, and render the therapy more nearly continuous. Continuous extracorporeal blood therapies also have a number of disadvantages in comparison with HD and PD. CVVH and CVVHD require considerable attention from nurses to ensure adequate blood flow, monitor anticoagulation status, adjust ultrafiltration rate, and calculate patient fluid balance. The patient is immobilized during the treatment. Continuous heparin administration increases risk of bleeding. Central venous or femoral catheters often provide insufficient blood flow and carry risks for infection and sepsis. In spite of the name and intent, the average duration of individual treatments with “continuous” blood therapies is 20 hours before clotting of the system or need to transport the patient for diagnostic or therapeutic procedures. Some centers previously performing CVVH or CVVHD for acute dialysis have begun using sustained low-efficiency dialysis with extended duration (SLED, 6–8 hours/day) to avoid many of the problems of CVVH or CVVHD, limiting heparin use and immobility of the patient while keeping the advantages of vascular stability and improved clearances. By contrast, PD is a truly continuous dialysis therapy associated with potentially fewer risks. It requires less nursing effort than CVVH, CVVHD, or SLED and allows more patient mobility during therapy.