Fluid management and risk factors for renal dysfunction in patients with severe sepsis and/or septic shock

In patients with severe sepsis and septic shock, acute renal failure (ARF) is an independent factor for mortality [1, 2]. In the last decade, significant efforts were made to standardize the treatment of septic shock [3, 4]. One of the most important recommendations is volume expansion that could also prevent ARF [5, 6]. However, the type of fluid, especially the use of colloids, for volume expansion in septic shock remains a matter of debate [7‐11]. Indeed, despite a larger plasma volume expansion power [12‐14], the use of hydroxyethylstarch (HES) is not related to better outcomes when compared to isotonic crystalloids. In addition, use of HES has been associated with the development of an impaired renal function [7, 8, 15]. Thus, some experts suggest avoiding the use of HES in septic shock [7‐9].

Because recent HES have a lower molecular weight (HES 130/0.4), some authors advocate their use in septic patients [16, 17]. In fact, controversial results have been reported in the literature [18, 19]. In the field of renal transplant, two studies suggested a better renal tolerance with low molecular weight HES, than with high molecular weight HES or gelatins [20, 21]. In patients with septic shock, a recent study suggested that even the use of low molecular weight HES was associated with renal dysfunction. In this study, HES 130/0.4 was used in the first phase of the study when the initial optimization of cardiac preload could be suspected as suboptimal when compared with the final phase using crystalloids [22].

In 2006, a regional program of quality improvement in patients with severe sepsis and septic shock was performed in 15 Southern French ICUs in the Languedoc Roussillon region. After an observational period (six months), an educational program was implemented to optimize the management of patients with severe sepsis and/or septic shock (Sepsi d'Oc study) [23]. As in other studies, the implementation of ten recommendations based on the Surviving Sepsis Campaign [3] and French recommendations were associated with an absolute reduction (13%) of 28-day mortality rates [23].

During this study, information on the type and the volume of fluid was collected in the first 24 hours of patient management. We hypothesized that, in severe sepsis or septic shock patients, the use of HES 130/0.4 may be associated with the development of renal dysfunction. The study was aimed at finding the variables in the first 24 hours that were associated with the occurrence of renal dysfunction in our cohort of patient. The secondary aims were to determine the volume of each type of fluid used in these patients, the relationship between the need for renal replacement therapy (RRT) and the type of fluid, and the outcome and the type of fluid.

The present study was a part of Sepsi d'Oc (grants from the University Hospital of Nîmes), a quality improvement program for the management of severe sepsis in 15 ICUs in southern France (Languedoc Roussillon, population: 2,402,000 habitants) [23]. Therefore, the Institutional Review Board at the Nîmes University Hospital approved this study and stated that informed consent was waived. The patients or next of kin were informed of the study and could decline participation in the study.

The present study focuses on the factors associated with the occurrence of renal dysfunction in patients with severe sepsis and septic shock. In our cohort, 73% of patients were given a combination of HES and crystalloids. With respect to renal dysfunction [27], male gender, a high SAPS II score, no decrease in SOFA scores, the case-mix (surgery), and the interventional period of the Sepsi d'Oc study were identified as risk factors for renal dysfunction. The factors associated with the need for RRT were the baseline value of plasma creatinine and the need for vasopressors. The administration of HES 130/0.4 in the first 24 hours of resuscitation was not associated with a risk of renal dysfunction.

The Sepsi d'Oc study [23] was aimed at improving the initial management (during the first 24 hours) of patients with recent severe sepsis and/or septic shock (< 24 hours). Based on the Surviving Sepsis Campaign [3], ten recommendations were implemented in 15 ICUs of the Southern French 'Languedoc Roussillon' region. This before-after study resulted in a 28 day-mortality reduction. This finding was concordant with those of recent, large studies [28‐32].

In the first 24 hours of management, more than 80% of the patients received HES. The volume of colloids was 830 ± 731 ml with a third quartile at 1,800 ml. The use of comparable volumes of HES in patients with severe sepsis or septic shock has already been reported elsewhere [19, 33]. In the present study, the volume of infused HES was in the range of recommended doses, as only 2% of patients received more than 50 ml/Kg of HES. Our findings are globally consistent with the available literature, suggesting that our population is well-representative of severe sepsis and septic shock patients. Of note, the total volume during the initial period of severe sepsis and or septic shock was lower than that reported by other groups [29, 34]. The occurrence of renal dysfunction and the need for RRT are associated with poorer outcomes, as previously reported [1, 2, 35].

In the Sepsi d'Oc study, the interventional period was associated with a larger occurrence of renal dysfunction but a higher 28-day survival rate. This finding suggests that an optimization of the initial management of patients with severe sepsis and septic shock may decrease the impact of renal dysfunction on the outcomes, as recently suggested by Badin et al.[36]. Moreover, the decrease in the 28-day mortality rate could expose more patients to the risk of renal dysfunction.

The present study failed to show that a low molecular weight HES is associated with poor renal outcomes. Initially, the potential deleterious effect of HES was demonstrated in the renal graft setting with previous non-low molecular weight types of HES [37, 38]. The suggested mechanism was an osmotic nephrosis. In ICU patients with septic shock, Schortgen et al.[39] also reported that more renal dysfunction was predominant in patients receiving HES, but this did not result in an increase in mortality or the need for RRT. In the VISEP study [15], a higher rate of renal dysfunction was reported but this did not impact the 90-day mortality rate. These previous studies used high molecular weight HES that were sometimes given in extremely large doses.

Some studies have shown that HES 130/0.4 is not associated with the development of impaired renal function [18, 19]. In the renal graft setting, two studies have shown that HES 130/0.4 is less deleterious in terms of renal function than the old generations of HES and gelatins [20, 21]. Elsewhere, in ICU patients, three studies reported that HES 130/0.4 administration was associated with poor renal outcome [8, 22, 40]. In the study reported by Bayer et al.[22], the baseline Central Venous Pressure (CVP) value was lower and more albumin and platelets and fresh frozen plasma were administrated in the HES group than in the gelatin and crystalloids groups. Moreover, patients in the HES group were more exposed to nephrotoxic product. These differences could explain why more renal dysfunctions were observed in this group without an increase in the need for RRT. In the study reported by Shortgen et al.[40], patients requiring HES administration previously received larger amounts of fluids. Moreover, they required more mechanical ventilation suggesting that they were more seriously ill whereas the SAPS II score was similar. In contrast, other groups reported similar findings to those of the present study [18, 19]. In the SOAP study [41], the infusion of HES was more frequent in the most seriously ill patients (higher baseline SAPS II and SOFA scores, more mechanically ventilated patients, more patients requiring blood products, more patients with severe sepsis and more patients with shock during their ICU stay) [18]. However, the multivariate analysis did not find that HES were associated with an increased risk of renal dysfunction. The study reported by Boussekey et al.[19] also failed to find a deleterious effect of HES in patients with severe sepsis and/or septic shock. Moreover, volumes of infused HES were similar to those reported in the present study (763 ± 595 ml on day 2, 1,031 ± 800 ml on day 7, 1,361 ± 1,393 ml on day 21). Our study suggests that in a real life situation, the physicians tend to respect the recommended doses of HES. This could probably partly explain the different results with previous studies in which HES was widely given. The present study clearly shows that the use of adequate dosages of HES is not associated with the occurrence of renal dysfunction.

The present study has several limitations. First, it focused on the use of different therapies in the first 24 hours of initial management of patients with severe sepsis and/or septic shock. We cannot report the real volume of fluid after 24 hours. However, Boussekey et al.[19] showed that the main dosage of HES is administrated in the first two days. Second, the present study was a cohort study and several biases could interfere with the findings. The physicians were not blinded, and a pre-selection of patients may have played a role in our results. However, this is a study reporting real-life practice, giving a partial response to an unresolved issue. One should note that more than 70% of our patients received both colloids and crystalloids, while randomized clinical trials often favor the use of a single type of fluid [15, 39]. Even if a multivariate analysis was performed, further randomized clinical trials will probably provide a definitive answer.

In the present study, renal dysfunction was independently associated with male gender, SAPS II score, a surgical patient, no decrease in SOFA score in the first 24 hours, and the interventional period of the Sepsi d'Oc study whereas RRT was independently associated with the need for vasopressors and the baseline value of serum creatinine. HES was widely used in patients with severe sepsis and septic shock in the initial 24-hour period but was not associated with deleterious impacts on renal function.