Abstract
Two articles on the controversial issue of colloid resuscitation feature this month in the pages of this journal. One is by Hunter et al., on the use of a hybrid colloid/crystalloid fluid resuscitation regime with the aim of mitigating ‘fluid creep’, 1 and the second which suggests that colloid resuscitation may impact on skin graft success. 2
In the former study, data on the use of a mixed albumin and crystalloids resuscitation strategy in burns are described as a hybrid regime that may mitigate ‘fluid creep’. 1 The authors introduce the concept of fluid creep and highlight the perceived effects of a Cochrane review from 1998.
Interestingly, in the ALBIOS trial, a multicentre, open-label randomised controlled trial which included over 1800 critically ill patients with severe sepsis, allocated to receive either a combination of 20% albumin and a crystalloid solution or a crystalloid solution only, there was no statistically significant difference in the total daily amount of administered fluid, the mortality and the other secondary outcomes, between the two arms. 3
The retrospective cohort study by Hunter at al. contains a relatively small patient population and is focused on the surrogate endpoint of volume of fluid administered, rather than more relevant, patient-centred endpoints such as mortality or morbidity. The authors acknowledge that both the ‘completeness of datasets’ and the ‘deviation from the protocol’ are potential limitations of the study. Nevertheless, the study does raise an interesting research question, suggesting a line of investigation possibly worth pursuing in the future. Whether the use of a mixed resuscitation regimen with albumin and crystalloids in the critically ill burns patient may mitigate fluid creep and improve harder outcomes, such as mortality and complications rates, should be further evaluated with large, methodologically robust, ideally international multi-centric randomised controlled trials.
The retrospective study by Isitt et al. 2 attempts to identify factors potentially associated with graft failure. The study includes a rather small patient sample and focuses primarily on the endpoint of graft failure, excluding patients who died. The non-inclusion of deceased patients in the study, together with its retrospective nature, raises important questions of methodology and bias, while the small patient sample means the study may have been simply underpowered and hence have failed to detect significant differences, even where those differences truly existed. Furthermore, all ‘colloids’ appear to be grouped into the same category: this is a significant weakness of the study, as colloids are a heterogeneous group, especially in terms of safety profile in the critically ill patient. In the 6S trial, which randomised 798 severely septic patients to hydroxyethyl starch 130/0.42 versus Ringer’s acetate, the colloid arm patients had increased 90-day mortality risk and were more likely to undergo renal replacement therapy. 4
In conclusion, both studies have provided interesting food for thought and debate for the readership, but more work needs to be done. The underlying issues highlighted require sufficiently large and methodologically rigorous trials, if we are to draw definite conclusions.