Perioperative goal-directed haemodynamic treatment and the equity of differing modalities.

Editor—We read the recent article written by Dr Bartha and colleagues addressing goal-directed haemodynamic treatment (GDHT) in patients with proximal femoral fracture with great interest. This subset of surgical patients is indeed elderly and at high risk with 20% 4 month mortality. 3 Therefore, an evidence base to guide clinicians in their perioperative care to optimize clinical outcome is very much required. However, the improvements we can make are heavily influenced by significant patient co-morbidities and also financial viability as the annual spend for this group is already above £385 million in the NHS. While we accept the authors’ point that there is no specific evidence to support an influence on postoperative complications, published work showing a reduced time of fitness to discharge does imply a reduction in morbidity. 5 We also feel that it is important to set this in the context of evidence to support the reduction in postoperative complications in other high-risk surgical groups. 7 Importantly, a stance has been taken by both NICE and the UK GIFTASUP guidelines in supporting the use of GDHT in high-risk surgical groups. Secondly, and most importantly, although the authors highlight the differences between their recent piece of work and previously published research, they do not emphasize the importance of comparing different technologies. This issue is brought to attention again in this study using a LiDCO generated stroke volume in an attempt to validate a GDHT model based on Shoemakers’ original work using a pulmonary artery (PA) catheter. As numerous less invasive technologies are now available, the use of the PA catheter has declined. However, it is not necessarily the case that these technologies are directly comparable either with the PA catheter or with each other. Until the monitors are individually validated against an existing standard and appropriate equivalent haemodynamic parameters and fluid protocols are defined, using the technologies interchangeably may only lead to confusion and a failure to demonstrate a benefit to patients. A question that remains unanswered is whether improved outcome is a result of optimizing oxygen delivery or a consequence of controlled fluid therapy? This issue is well illustrated in a recent publication in the BJA comparing stroke volume optimization between the oesophageal Doppler monitor (ODM) and the uncalibrated LiDCOrapid in elective high-risk colorectal surgery patients. The ability of the LiDCOrapid to track changes in the stroke volume measurements of the ODM was weak with a calculated concordance of 80%. The sensitivity and specificity to detect a positive fluid challenge was 48% and 80%, respectively. We note that the authors used calibrated LiDCO but what is not clarified is whether the monitor was recalibrated after establishing spinal anaesthesia or after vasopressor therapy since the algorithm is sensitive to changes in systemic vascular resistance and measured oxygen delivery may not have remained accurate. It is interesting to note that the ODM itself, although considered a gold standard for perioperative monitoring in the UK, may have significant limitations. It has been shown that only 37% of patients who have a decrease in SV perioperatively respond to a subsequent fluid bolus and also absolute measurements of ODM stroke volume can also vary from true values by 40%. Returning to this trial, it is unfortunate that recruitment remained less than planned or required to provide statistically robust results for the short-term primary and secondary outcome measures; however, we suspect that even if the required number of patients were evaluated, the results from this study may not have been comparable with previous studies using different technologies and assuming them to be providing interchangeable data. While we appreciate the work involved and the contribution that this study has made to progressing research in this area, we feel that it is crucial to validate and compare these technologies and then apply them consistently in order to be able to reap the benefits that they offer our patients. The question that could be answered is whether outcomes remain the same in spite of the monitor and optimization protocol used.