Hypophosphatemia during continuous veno-venous hemofiltration is associated with mortality in critically ill patients with acute kidney injury

Hypophosphatemia is one of the frequently encountered electrolyte disorders in critically ill patients, with a prevalence ranging from 20% to 40% [1‐4] and even reaching 80% in septic patients [5]. Because the common mechanism in hypophosphatemia-caused complications is impaired energy metabolism, hypophosphatemia has also been described as a metabolic disturbance leading to cellular dysfunction in multiple organ systems [6]. Therefore, hypophosphatemia may lead to some serious clinical problems, such as respiratory muscle dysfunction [7, 8], decreased myocardial contractility [9], ventricular tachycardia [10] and neuromuscular disturbances [11, 12]. Many studies have shown an association between hypophosphatemia and increased mortality [13‐15]. However, it remains unclear whether hypophosphatemia actually contributes to mortality or is merely a marker for severity of illness.

Three main mechanisms lead to hypophosphatemia: decreased intestinal absorption, increased renal excretion and internal redistribution of inorganic phosphate. The main causes of hypophosphatemia in critically ill patients include severe infection, trauma, postoperative state, malnutrition, respiratory alkalosis and diabetic ketoacidosis [6]. Renal replacement therapy may also lead to hypophosphatemia, particularly for a prolonged period [16‐18]. The incidence of hypophosphatemia was reported to range from 54.0% to 65.1% in the Randomized Evaluation of Normal vs. Augmented Level (RENAL) Replacement Therapy Trial and from 10.9% to 17.6% in the VA/NIH Acute Renal Failure Trial Network (ATN) Study [19, 20]. Because renal replacement therapy is one of the most widely used methods to treat critical illness, especially acute kidney injury (AKI), it is crucial to study the effects of high-incidence complications on survival, including hypophosphatemia. Unfortunately, few epidemiologic data about renal replacement therapy associated hypophosphatemia and global clinical outcome have been reported to date.

In the present study, we retrospectively investigated a critically ill patient cohort with AKI who received continuous veno-venous hemofiltration (CVVH) therapy to determine whether hypophosphatemia during CVVH is associated with mortality.

This study was approved by the Institutional Ethics Committee of Zhejiang University, and informed consent was obtained from the patients and/or their guardians.

The baseline demographics and clinical characteristics of the patients are summarized in Table 1. The mean age (±SD) for patients of the study cohort was 59.72 ± 17.79 years, and 498 patients (65.5%) were male. A total of 406 patients (53.4%) were diagnosed with sepsis according to the criteria published by the International Sepsis Definitions Conference [22]. The mean APACHE II score (±SD) at the initiation of CVVH therapy was 21.04 ± 6.64. A total of 471 patients (62.0%) were injected with sodium glycerophosphate according the manufacturer’s drug instructions. A total of 521 patients (68.6%) had hypophosphatemia episodes during CVVH therapy, and 109 patients (14.3%) developed severe hypophosphatemia according to our definition. A total of 387 patients (50.9%) were alive at day 28 after initiation of CVVH. Univariate analysis indicated that survivors were significantly different from nonsurvivors with regard to some demographic and clinical characteristics, including younger age, less likely to have chronic ill health (as defined by the APACHE II criteria), lower occurrence of sepsis, lower APACHE II scores and lower occurrence of severe hypophosphatemia (11.4% vs. 17.4%; P = 0.022). However, gender, recent surgery, median length of mechanical ventilation after initiation of CVVH therapy, median dose and therapeutic time of CVVH, the proportion of patients receiving phosphate supplementation and the occurrence of hypophosphatemia (67.2% vs. 70.0%; P = 0.435) demonstrated no significance between the survivor and nonsurvivor cohorts. Univariate factors with P-values less than 0.2 included age, chronic ill health, sepsis, APACHE II score and presence or absence of severe hypophosphatemia episode during CVVH therapy. These patients were included in the Cox proportional hazards survival model to avoid the potential influence of some confounding risk factors. The multivariate analysis indicated that the occurrence of severe hypophosphatemia episode during CVVH therapy was dependent on outcome (P = 0.700), whereas APACHE II score was independent of 28-day mortality (Figure 1 and Table 2).

Thereafter we analyzed the subcohort of 521 patients who had developed hypophosphatemia during the CVVH therapy period. In this cohort, the mean ratio of CVVH therapy days with hypophosphatemia to total CVVH therapy days was 0.58. Therefore, the ratio ≥0.58 was defined as a high ratio, and the ratio <0.58 was defined as a low ratio. A total of 260 patients (49.9%) were alive at day 28 after initiation of CVVH of this cohort, and the univariate analysis indicated that survivors were significantly different from nonsurvivors regarding some factors, including younger age, less likelihood of chronic ill health, a lower occurrence of sepsis, lower APACHE II score and lower occurrence of high ratio of CVVH therapy days with hypophosphatemia to total CVVH therapy days (36.9% vs. 55.9%; P < 0.001) (Table 3). However, gender, recent surgery, median length of mechanical ventilation after initiation of CVVH therapy, median dose and therapeutic time of CVVH and proportion of patients receiving phosphate supplementation demonstrated no significance between the survivor and nonsurvivor cohorts (Table 3). Univariate factors with P-values less than 0.2, including age, chronic ill health, sepsis, APACHE II score and high or low ratio of CVVH therapy days with hypophosphatemia to total CVVH therapy days were recruited in the Cox proportional hazards survival model and indicated that the ratio of CVVH therapy days with hypophosphatemia to total CVVH therapy days was independently associated with mortality. Compared with the patients with a low ratio, those with a high ratio conferred a 1.451-fold increase in 28-day mortality rate (95% CI = 1.103 to 1.910; P = 0.008) (Figure 2 and Table 4). Meanwhile, APACHE II score was also an independent factor associated with 28-day mortality.

AKI is a frequent problem in the ICU associated with substantial morbidity, mortality and healthcare expenditures. Severe AKI is always accompanied by dysfunction in other organs and is characterized by high mortality. Renal replacement therapy remains the cornerstone of supportive care [23]. To date, some important issues regarding renal replacement therapy are still unclear, and any progress regarding interventions may contribute much to the global outcomes of these critically ill patients. Therefore, in our present study, patients diagnosed with AKI receiving CVVH therapy were recruited as the study cohort to determine the relationship of clinical outcomes and hypophosphatemia, one of the frequently complications of renal replacement therapy.

It has been indicated that, besides CVVH therapy, the main causes of hypophosphatemia in critically ill patients may include severe infection, trauma, postoperative state and malnutrition [6]. Indeed, in the present study, 272 patients (35.8%) had received surgical interventions and 406 patients (53.4%) were diagnosed with sepsis. Therefore, hypophosphatemia may occur in some patients before the initiation of CVVH therapy because of the illness itself. Importantly, the potential that hypophosphatemia could be a surrogate marker for something else can never be excluded, even during the CVVH therapy period.

It was reported in the ATN Study that the incidence of hypophosphatemia was 10.9% in the low-intensity group and 17.6% in the high-intensity group [20]. In the RENAL Trial, the incidence of hypophosphatemia in the low- and high-intensity groups was 54.0% and 65.1%, respectively [19]. The difference in hypophosphatemia incidence may due to the following causes. The modality of renal replacement therapy performed in the ATN Study included intermittent hemodialysis, sustained low-efficiency dialysis and continuous veno-venous hemodiafiltration. In the RENAL Study only continuous veno-venous hemodiafiltration was administered. The therapeutic dose of renal replacement therapy adopted in the ATN Study was lower than that in the RENAL Trial (35 ml/kg/h vs. 40 ml/kg/h in the high-intensity group). The results of comparison of the two study implied that phosphate clearance on continuous veno-venous hemodiafiltration might be significantly greater than intermittent hemodialysis because of ongoing intercompartmental mass transfer and larger filter pore size. In the present study, the modality of renal replacement therapy was CVVH, the mean therapeutic dosage was closer to that used in the RENAL Trial design, and the incidence of CVVH-associated hypophosphatemia according to our data was also similar to the RENAL Trial.

Hypophosphatemia can be caused by decreasing absorption, increasing renal excretion or internal redistribution of inorganic phosphate. In critically ill patients, inorganic phosphate redistribution across the cell membrane is a common phenomenon [24]. Therefore, serum phosphorus levels do not accurately reflect total body phosphorus stores, and this bias can be aggravated by CVVH therapy because of the quick serum phosphorus clearance during the intervention. The ratio of CVVH therapy days with hypophosphatemia to total CVVH therapy days was calculated in the present study to avoid the bias of the serum phosphorus test. The present study showed that a single test of serum phosphorus concentration could not serve as an independent factor predictive of mortality even if it reached the standard of severe hypophosphatemia. Interestingly, the ratio of CVVH therapy days with hypophosphatemia to total CVVH therapy days was proved to be a factor independent of the global clinical outcome through the multivariate analysis, indicating that a high ratio conferred a higher 28-day mortality rate. The results suggested that the persistence of hypophosphatemia had the potential to harm critically ill patients. Once hypophosphatemia happens, the therapeutic strategy might be adjusted, including more phosphate supplementation as well as regulation of renal replacement dosage or modality, to avoid persistence of hypophosphatemia.

To date, whether the impact of intensity or dosage of renal replacement therapy is crucial to the outcomes of critically ill patients is controversial. Despite the benefits of high-intensity renal replacement therapy, including superior hemodynamic stability, metabolic clearance and volume control, no survival advantage can be demonstrated with its use [25]. Although study design may explain much of this paradox, it is also quite plausible that the complications of the high-intensity renal replacement therapy offset its potential benefits in critically ill patients. Unfortunately, some outcome–dose studies have not even discussed adverse effects. In our data, normal phosphate supplementation treatment failed to reverse hypophosphatemia during CVVH because the majority of the patients with hypophosphatemia (68.6%) received regular sodium glycerophosphate injections. Thus, a new strategy should be adopted to resolve this clinical problem. The present study highlights the need for analysis of adverse effects and clinical strategies to correct hypophosphatemia during CVVH therapy, including phosphate supplementation, especially the use of phosphate-containing replacement fluid to decrease the high mortality rate of severe AKI patients [17, 18, 26].

The present study has several limitations. The retrospective study design decreased the power of the conclusions. As the endpoints we adopted, we dealt with the 28-day mortality in the present study, which was only a surrogate outcome. In addition, various clinical settings may lead to hypophosphatemia in critically ill patients, so we could never exclude the potential that hypophosphatemia could be a surrogate marker for something other than CVVH. A prospective study with various days to mortality may resolve this bias.

● Hypophosphatemia during CVVH was associated with the global clinical outcome of critically ill patients with AKI.

● The ratio of CVVH therapy days with hypophosphatemia to total CVVH therapy days was independently associated with 28-day mortality.

● A high ratio of CVVH therapy days with hypophosphatemia to total CVVH therapy days conferred a higher mortality rate among critically ill patients with AKI.