Criteria abstracted from The Users' Guide to Medical Literature, from the Health Information Research Unit and Clinical Epidemiology and Biostatistics, McMaster University

Highlighted lines and questions below provide links to the pertinent description of criteria in The EBM User's Guide, now available at the Canadian Centres for Health Evidence

Article Reviewed:

Please visit the new Evidence Based Journal Club Reviews

Daily hemodialysis and the outcome of acute renal failure

Schiffl H, Lang SM, Fischer R..

N Engl J Med 2002; 346:305-10. [abstract; full-text for subscribers]

Reviewed by Xiomara Garcia, MD, Pediatric Critical Care Section, Department of Pediatrics, Baylor College of Medicine, Houston, Texas. Review posted June 6, 2002

N.B.: An "expression of concern" regarding the scientific conduct in this trial was published in the May 22, 2003 issue of the New England Journal of Medicine..

---

I. What is being studied?:

The study objective:

Determine the effect of daily intermittent hemodialysis compared with conventional (alternate-day) intermittent hemodialysis, on survival among patients with acute renal failure.

The study design:

Prospective, controlled, unblinded, single institution study.

The patients included:

Adults with acute renal failure admitted to Medical and Surgical Intensive Care Units at the University Hospitals of Munich, Innenstadt, Germany. Main inclusion criteria: diagnosis of severe acute tubular necrosis (ATN) caused by a recent ischemic or nephrotoxic injury with an anticipated need for intermittent hemodialysis for a least 1 week. Severe ATN was defined as a rapid increase in serum creatinine of 1 mg/dl or marked azotemia (serum creatinine 4 mg/dl) and a history of prolonged and profound hypotension, a severe overdose of nephrotoxins and the presence of risk factors for nephrotoxic acute tubular necrosis, or an excessive body burden of endogenous nephrotoxic pigments.

The patients excluded:

Those with functional azotemia, urinary tract obstruction, acute interstitial nephritis, rapidly progressive glomerulonephritis, chronic renal failure ( serum creatinine > 3 mg/dl), renal transplantation, previous hemodialytic or filtrative therapy for the same episode of renal failure, and indications for continuous renal replacement (CVVH-D) were excluded.

The interventions compared:

Daily intermittent hemodialyisis was compared with conventional (alternate-day) intermittent hemodialysis. The decision to begin or terminate hemodialysis was made by the attending nephrologist in the ICU.

The outcomes evaluated:

The primary end point of the study was survival 14 days after last session of hemodialysis. The secondary end points were: duration of acute renal failure and frequency of therapy-related complications.

II. Are the results of the study valid?

Primary questions:

1. Was the assignment of patients to treatments randomized?

No, consecutively enrolled patients were assigned in alternating order to receive daily or conventional hemodialysis. To minimize bias, once the nephrologist recommended hemodialysis and the patient was enrolled, treatment was assigned without knowing identity of patient or clinical characteristics that were predictors of the outcome. This method of randomization by alternating patient assignment does leave some potential for bias. For example, if two patients were ready for enrollment at the same time and everyone was aware of the group assignment of the last patient enrolled, the order of enrollment could be manipulated to assure assignment to one group verses the other.

2. Were all patients who entered the trial properly accounted for and attributed at its conclusion?

Was followup complete?

Yes, a total of 160 patients were assigned to the two treatment regimens in alternating order. 14 patients were withdrawn during the course of the study, before the final data analysis. In the daily dialysis group, two had biopsy-proven, rapidly progressive glomerulonephritis (Good pasture's disease), two were switched to continuous renal replacement therapy because of clinical deterioration, and two required surgery during the first week of the study. In the conventional dialysis group, four were switched to continuous renal replacement therapy because of clinical deterioration, and two required surgery. Of the remaining 146 patients, 72 patients received hemodialysis every other day, and 74 received daily treatment. The baseline characteristics of the two treatment groups were similar.

Were patients analyzed in the groups to which they were randomized?

Yes. An intention-to-treat analysis was performed to accommodate the 14 patients removed from the study. Efficacy analysis was also performed and included only those patients who completed the trial. The patients that were non-eligible when analyzed according to the intention to treat approach were kept in their respective treatment groups. The analysis of recovery of renal function included survivors and not survivors because a few of the non-survivors had a partial recovery of renal function before they died.

Secondary questions:

3. Were patients, health workers, and study personnel "blind" to treatment?

No. Physicians, nurses, patients were unaware of treatment assignment until first session completed. Because of the nature of the therapy (daily verses alternating days), after day one of hemodialysis, blinding of nurses and physicians was not maintained.

4. Were the groups similar at the start of the trial?

Yes. Almost, except for the fact that there were more patients in the Medical intensive care unit than in the Surgical intensive care unit. Also the BUN in the group of conventional hemodialysis started at higher level than the daily hemodialysis group. However, these weren't statistically significant differences, and it is unlikely they were clinically significant differences.

5. Aside from the experimental intervention, were the groups treated equally?

It is not clear. There are not specifications about other interventions or treatments (except nutritional support) provided to the patients enrolled in the study.

The mean duration of hemodialysis session, the average blood flow-rate, and the mean prescribed and delivered "doses" did not differ significantly between the groups. Obviously, the daily treatment group received nearly twice the "dose" of dialysis per week. The lack of blinding allows for the possibility that patients were unconsciously treated differently in the two groups.

III. What were the results?

1. How large was the treatment effect?

All data were reported as mean and standard deviation. Also odds ratios for death according to selected variables from the multiple logistic regression analysis were reported Using intention-to-treat analysis, mortality in the daily group was 28% compared to 46% in the intermittent group (p=0.01). By efficacy analysis (excluding drop-out patients): 26% mortality in the daily group compared to 43% mortality in the intermittent group (p=0.04).

Treatment	Non Survival	Survival	Total patients
Daily hemodialysis	22 (0.28)	58 (0.72)	80
Conventional hemodialysis	37 (0.46)	43 (0.57)	80

*intention to treat numbers

ARR = 0.46 - 0.28 = 0.18 (95% CI 0.04, 0.32)
RR = 0.28/0.46 = 0.61 (95% CI 0.47, 0.75)
RRR = 1 - 0.61 = 0.39 (95% CI 0.25, 0.53 ) Daily hemodialysis reduced the relative risk of mortality by 39%.
NNT = 1/0.18 = 5.6 (95% CI 3, 25) Approximately 6 patients need to be treated with daily hemodialysis instead of conventional intermittent hemodialysis to prevent one death

2. How precise was the estimate of the treatment effect?

The absolute risk reduction is 18% with a 95% CI of 4% to 32%. The difference in the risk of death between patients with daily hemodialysis and those who had conventional hemodialysis will range between 4 to 32%. The number needed to treat with daily hemodialysis instead of conventional hemodialysis to prevent one death, could be as few as 3 patients or as high as 25 patients. These are not terribly precise confidence intervals but certainly they confirm that the point estimate is significant.

Multiple logistic regression analysis demonstrated that four of the tested variables had a significant effect on survival. A greater severity of illness and the presence of sepsis at the time of enrollment were negatively correlated with survival. Normal urinary output at the time of enrollment (OR for death related to oliguria vs normal urinary output is 3.02 (95% CI 1.35-6.77, p < 0.007) and assignment to the regimen of daily hemodialysis were positively correlated with survival. The odds ratio for death in the conventional group vs. the daily group, adjusting for all cofounders, was 3.92 (95% CI 1.68, 9.18). This is a statistical and highly clinically significant result.

IV. Will the results help me in caring for my patients?

1. Can the results be applied to my patient care?

The majority of the patients in our Pediatric Intensive Care Unit who are seriously ill with acute renal failure often require CVVH-D secondary to blood pressure instability. Although they may be "fluid overloaded" they often have capillary leak syndrome, so intravascularly they are fluid depleted; they may not be able to tolerate conventional or even daily hemodialysis. But for stable patients, these results might apply.

The type of dialysis membranes used during hemodialysis were not specified (this information is relevant because the biocompatibility of the membrane may have affected the outcome).

No criteria for the use of continuous regimen rather than an intermittent regimen were provided. We do not know the number of patients who were enrolled in the study in relation to the number of patients who were treated with continuous renal replacement and also it would be interesting to know the mortality in this group of patients compared with the two study groups.

2. Were all clinically important outcomes considered?

Yes, in terms of clinical parameters of recovery and most importantly, mortality, all important outcomes were considered. They measured survival and also related complications as: systemic inflammatory response syndrome or sepsis, respiratory failure, changes in mental status, gastrointestinal bleeding as well as recovery of renal function.

3. Are the likely treatment benefits worth the potential harms and costs?

I think that the results of this study can be applied to patients with acute renal failure and a clinical diagnosis of acute tubular necrosis who, at the start of therapy are not so severely ill that a continuous regimen is warranted. I think that daily hemodialysis is associated with less dramatic changes in plasma components, a reduced need for fluid removal and hence less severe hemodynamic instability. As a result there may be a lower incidence of progression to oliguria and recovery could be more rapid. Also, daily exposure to the dialysis membrane results may result in more efficient filtration of blood components that contribute to adverse systemic responses. Another study to consider these issues as outcomes would be indeed be useful.

To some degree, the answer to this question may also depend on an institution's resources. Are sufficient equipment and personnel available to offer daily hemodialysis for patients? If resources are adequate, given the low NNT, daily hemodialysis seems to be a very reasonable and beneficial alternative to intermittent hemodialysis. Whether this is applicable to pediatric patients remains to be seen.

Comments

Back to the EB Journal Club Index