The data collected in this study show that in adult patients with severe malaria CVP is not a reliable predictor of volume status, nor is it a clinically useful measure of the likelihood of developing pulmonary oedema, and therefore cannot be used as a guide for fluid therapy. The data collected in this study do not support the WHO recommendation that a CVP of 0-5 cm H2O should be used as a resuscitation target; the vast majority of all patients with a CVP recorded in the target range had concurrent hypovolaemia, pulmonary oedema or both while their CVP was in the target range. There was no correlation between the baseline CVP and the likelihood of the cardiac index (CI) being volume responsive, nor between the change in CVP (ΔCVP) and change in CI with fluid loading. While ΔCVP was correlated with the change in lung water with fluid loading there was such a large overlap in values that CVP is rendered clinically useless as a predictor in this situation. There was no correlation between CVP and renal function.
There is no literature on the use of CVP as a resuscitation target in patients with severe malaria, however pulmonary oedema has been shown not to be correlated with either elevated CVP or pulmonary artery occlusion pressure in previous studies of this group of patients [21, 22]. In the only large study to examine their relationship, measurements of CVP were actually lower in patients with pulmonary oedema than those in patients without this complication .
In recent years, there has been a reduction in the emphasis on pressure-based measures as endpoints for resuscitation in critically ill patients [24–27]. CVP is not a good predictor of fluid responsiveness in either healthy volunteers  or hypovolaemic critically ill patients , and current opinion in well equipped intensive care units is that CVP measurement alone should not be used to define the state of ventricular filling (preload) or the potential of patients to respond to a fluid challenge [15, 16].
Adherence to the bundle of care recommended in the Surviving Sepsis Guidelines has been shown to improve the outcomes of patients suffering from severe sepsis. Fluid resuscitation to a CVP of ≥ 8 mmHg in patients with persistent hypotension is one of the key tenets of these guidelines. However, when different components of the bundle of care are independently assessed investigators have shown that there was no association between achieving this CVP target and outcome [28, 29]. This has led some authors to question the appropriateness of a CVP target in guidelines .
Insertion of CVCs is not without risks: mechanical complications occur in 5-19% of patients, infectious complications occur in 5-26%, and thrombotic complications occur in 2-26% . Patients with severe malaria have thrombocytopenia and sometimes coagulopathy increasing the risk of haemorrhage. CVCs are relatively expensive, require medical and nursing expertise to insert and maintain and supporting radiology and microbiology services; all of which will often be lacking in the resource-poor settings where most patients with severe malaria are managed. In our series 25% of the patients had a significant complication of CVC insertion - a relatively high figure, but indicative of the resource-poor setting in which the patients were managed.
Central venous access for inotrope therapy is needed in only a minority of patients, since shock is a rare complication in severe malaria . Also, parenteral feeding in patients with coma is seldom indicated, since coma recovery is generally within 72 hours .
In our study the JVP was strongly correlated with CVP, a relationship that has been noted in other studies [8, 31] although the fact that in this study the JVP and CVP were measured by the same examiners concurrently potentially confounds this observation. However, there was similarly little correlation between the JVP and measures of volume status, pulmonary oedema or cardiac index or volume responsiveness. Furthermore JVP is a notoriously challenging sign to interpret, considerable variation exists in expertise in the assessment of JVP, and some studies have shown poor reliability of such assessments in critically ill patients [32, 33].
Ultimately, the goal of fluid therapy with the optimization of blood volume is to improve cardiac output and hence oxygen delivery to tissues. However the central pathophysiological mechanism of severe falciparum malaria is the disturbance of microcirculatory blood flow. The adherence of parasitized red blood cells to the endothelium - compounded by auto-agglutination, rosetting and reduced deformability of the red cells - leads to decreased tissue perfusion and the lactic acidosis which has repeatedly been shown to be the strongest predictor of mortality. Microvascular measures like the proportion of obstructed capillaries in rectal mucosa or estimates of the sequestered parasite biomass, have been shown to be strongly correlated with lactic acidosis and outcome in patients with severe malaria [34, 35]. Fluid loading to systemic haemodynamic targets may do little to overcome this microcirculatory obstruction whereas pulmonary oedema associated with fluid loading in malaria is common and has a mortality rate in adults in resource-poor settings of up to 80%. Recent data collected in African children with severe malaria have demonstrated increased mortality in children treated with aggressive fluid loading .
PiCCO derived volume based measures were more reliable as predictors of fluid responsiveness in this group of patients, as has been reported previously [37–40] However, the required expertise and resources limit its feasibility for deployment in the field. Moreover management of critically ill patients using PiCCO based measures has not yet been shown to improve outcome when compared to pressure based systems .
There are some important limits to the study. Only a small number of patients were enrolled and PiCCO derived measures were used as our gold standard measure of CI, GEDI and EVLW. The PiCCO system has not been validated in patients with severe malaria; it is theoretically possible that the microcirculatory derangement seen in malaria might affect the Stewart-Hamilton algorithm. The correlation seen between PiCCO measures may be argued to be the result of mathematical coupling although previous research has suggested that this is not the case . The study's findings do not preclude the possible utility of central venous access in some patients with severe malaria: central venous access would have value in patients requiring inotropic support or in whom peripheral access is not possible.
In summary, the data collected in this study show that invasive CVP measurement in adult patients with severe malaria provides little clinically useful information, while carrying appreciable morbidity and significant costs relevant to the resource-poor setting in which most patients are managed. In light of these data the WHO recommendations for the routine use of central venous lines - and a target CVP of 0-5 cm H2O - in patients with severe malaria should be reconsidered.