Intraoperative fluid management guided by oesophageal Doppler monitoringBMJ 2011; 342 doi: https://doi.org/10.1136/bmj.d3016 (Published 24 May 2011) Cite this as: BMJ 2011;342:d3016
All rapid responses
I am concerned at a couple of points with this paper. It is presented as research which it clearly is not.
To my eye the groups are far from matched. The "control" group are older 43.6% over 71 vs 37.8% over 71, they are sicker, ASA score 3 or greater in 41.7% vs 35% and there is more emergency surgery, 30.5% vs 27.3%. The authors state that as the POSSUM scores were similar this is not relevant. I am not so sure.
In my hospital there are many drivers to decrease length of stay and enhanced recovery programmes in most specialities. Length of stay is dropping all the time. This confounds the use of length of stay as a marker of better outcome.
I have little doubt there is a role for perioperative haemodynamic optimisation for some surgical groups but I do not think this article has helped take this forward a great deal.
Competing interests: No competing interests
We thank Dr Ashworth for his interest in our paper.
We agree that changes in perfusion are reflected by changes in the oxygen
saturation in the effluent of the circulation. Whether this can be used
for peri-operative fluid optimization is debatable with a recent trial
showing no improvement over standard care (1). Central venous
catheterization is expensive and invasive with well known associated
risks. Our quality improvement trial showed that anaesthetists were less
likely to use CVC when offered a better, less invasive alternative. We
think CVC monitoring should be reserved for patients requiring vasoactive
infusions, and not utilized for routine haemodynamic monitoring.
We thank Dr Wilson for his interest.
There have been numerous randomized controlled trials of Doppler. The aim
of this study was not to replicate such trials, but to explore barriers to
purchase and implementation of the Doppler and whether the benefits
predicted by research could be obtained in the real world. The sites were
deliberately chosen to represent the diversity of NHS providers and are
not all large centres. One site had no previous experience of Doppler use
intra-operatively, and across all sites the baseline usage in eligible
patients was only 12%. In this quality improvement project, the effect of
implementing recommended best practice (2) into routine service was
evaluated by comparing a consecutive cohort following implementation to a
historical cohort. There was no blinding or randomisation. The project
was approved by each Trust's clinical audit process and endorsed by
executive teams. Significant barriers to purchase and implementation were
overcome on all 3 sites as discussed in the report. A highly significant
increase in Doppler usage was achieved to 63% of eligible operations. This
was associated with a reduction in total and postoperative length of stay,
suggesting that the benefits predicted by research can be mirrored in
practice. The median length of stay across all sites was reduced from 12
to 10 days. Complications and in-hospital mortality are presented in Table
5, and fell non-significantly. The study design cannot exclude confounding
by other factors changing over time and potential for bias is discussed in
the paper. This quality improvement design could and should not replace
the type of randomized controlled trials conducted by Dr Wilson and
others. However we believe that our implementation study should aid and
encourage clinicians to translate research evidence into real benefit for
patients. The oesophageal Doppler now has evidence spanning the spectrum
from basic science through clinical trials to our real world quality
We thank Dr Miura and colleagues for their interest.
We implemented into routine practice a recommended intervention with a
strong evidence base (2). Differences in the group matching were
acknowledged in the paper. POSSUM is the best weighted, validated scoring
system for evaluating perioperative risk and has been used in many of the
RCT's of peri-operative fluid therapy. The use of different colloids at
different hospitals reflects variation in fluid prescribing practice
across the UK and increases the generalizability of the results. The
concern raised about colloid pharmacology is not supported by evidence
that differences in colloid molecular structure do not significantly
affect volume expansion or the accuracy of cardiovascular measurements
We note the commentator's active involvement with the NHS Networks
platform. The NHS Technology Adoption Centre has a specific remit and
funding for the purpose of supporting the adoption of technologies which
could be of significant benefit yet are underutilized. We suggest the
relevant peri-operative NHS Networks should encourage the implementation
of Doppler monitoring in accordance with NTAC's aims and with recent NICE
guidance (4), thus improving outcomes for major surgical patients.
1. Jammer I, Ulvik A, Erichsen C, L?demel O, Ostgaard G. Does central
venous oxygen saturation-directed fluid therapy affect postoperative
morbidity after colorectal surgery? A randomized assessor-blinded
controlled trial. Anesthesiology. 2010;113(5):1072-80
2. Powell-Tuck J, Gosling P, Lobo DN, Allison SP, Carlson GL, Gore M,
et al. British Consensus Guidelines on Intravenous Fluid Therapy for Adult
Surgical Patients - GIFTASUP. 2008. Available from:
http://journal.ics.ac.uk/pdf/1001013.pdf (Accessed 18 June 2011).
3. Lobo DN, Stanga Z, Aloysius MM, Wicks C, Nunes QM, Ingram KL,
Risch L, Allison SP Effect of volume loading with 1 liter intravenous
infusions of 0.9% saline, 4% succinylated gelatine (Gelofusine) and 6%
hydroxyethyl starch (Voluven) on blood volume and endocrine responses: a
randomized, three-way crossover study in healthy volunteers Crit Care
Med. 2010 Feb;38(2):464-70
4. MTG3 CardioQ-ODM (oesophageal Doppler monitor): NICE guidance
2011 Available from
http://www.nice.org.uk/nicemedia/live/13312/52624/52624.pdf (Accessed 18
Competing interests: DHC received expenses to travel to a meeting from Deltex Medical, this was after acceptance of publication in March 2011.
We are disappointed to read of this 'non-randomised "before and
after" project' that merely reproduces the randomised trial outcomes of
Wakeling et al (1) and Noblett et al (2) amongst others. There would
appear to be little sense in creating a further study that does not
advance our knowledge of goal-directed fluid therapy in major surgery.
Despite this study being headed under the umbrella of 'quality
improvement', we are surprised that there is no mention of advice (if not
approval) sought from the Research Ethics Committee. The assumption is
therefore made that this effort is actually 'service evaluation'?
According to the pre-operative patient data given, there must be
doubts raised as to whether the control and intervention groups are
comparable. The intervention group seemingly contains a much higher
number of younger patients (<60) who possess less morbidity (almost 10%
more ASA 3 in control). The project group does, in fairness, recognise
this difference despite the matching process, and utilises POSSUM scoring
in order to justify outcomes. Perhaps statistics of the group
demographics could be revealed? It is difficult to consider that these
groups are comparable on the basis of a single physiological scoring
We note the inconsistent choice of colloid solution across sites -
gelatin-based at Derby and Whittington versus starch-based at Manchester.
Can observations of cardiac output be comparable when the two have
Incidentally an editing error is noted in Table 4 regarding length of
stay - we are certain that all 658 patients did not require critical care
We have taken the liberty of perusing the web-based guide that has
been generated as a resource - the 'How to why to Guide' under NTAC. To
quote one of the 'care delivery principles':
"The effectiveness of Doppler-guided fluid management in reducing
hospital stay by an average of 3 days has not resulted in an increase in
readmission or reoperation rates. In fact, observational data from our
implementation sites has shown a 25% reduction in reoperation rates and
33% reduction in readmission rates when compared to our historical control
If the re-admission and re-operation figures are not statistically
significant (Table 5), then the statements that have been created are
certainly very bold.
We are curious as to what level of resource was expended in order to
construct the trademark 'How to why to Guide', let alone the NHS
Technology Adoption Centre - when a similar resource may have been
produced using the increasingly popular NHS Networks platform, which is a
free resource that was specifically developed to promote the sharing of
ideas and pooling of experience amongst NHS employees.
To conclude, it seems a shame that much effort and resource has been
expended on this 'implementation project', which cites the evidence-base
of no less than eight RCTs, and yet produces outcomes that cannot be
justified by the specific data presented.
1. Wakeling HG, McFall MR, Jenkins CS, Woods WG, Miles WF, Barclay
GR, et al.
Intraoperative oesophageal Doppler guided fluid management shortens
hospital stay after major bowel surgery. Br J Anaesth 2005;95:634-42
2. Noblett SE, Snowden CP, Shenton BK, Horgan AF. Randomized clinical
the effect of Doppler-optimized fluid management on outcome after elective
resection. Br J Surg 2006;93:1069-76.
Competing interests: No competing interests
I will declare my interest at the start, I have long championed so-
called "haemodynamic optimisation" in high-risk surgical patients,
starting with a randomised controlled trial published in this very journal
in 1999. I was also responsible for introducing oesophageal Doppler
monitoring as a standard for major surgery in my own hospital in 2000, and
I co-ordinated the change over to an alternative technology strategy 3
years ago, at the behest of colleagues who were less happy with using the
Doppler than me. I have used oesophageal Doppler many hundreds of times,
and still do on occasion.
I therefore read the paper by Kuper and colleagues with interest, but
have been with left with a degree of scepticism, as I really don't know
what to believe from this piece of work, which seems to be a form of
audit, yet with selective results presented in a research style and served
up to the audience as quality improvement. This matters greatly, as the
company that sells the device will undoubtedly be using this piece of work
in their marketing strategy, and the unwary will be influenced
There are many unanswered questions in this paper, apart from the
obvious huge potential for bias in the structure of this evaluation:
Why was a consecutive intervention cohort not compared to a
consecutive historical cohort? The authors chose matched controls instead,
this is not comparing like for like. How were these controls chosen? Was
there any blinding involved? Is this not beginning to sound like it should
have been a formal case-controlled study with ethical approval?
Why did the NICE Technology Adoption Centre not choose sites that had
no experience of using the oesophageal Doppler? The three sites chosen are
all large centres with enthusiasts driving change, why not include some
other general hospitals with no previous interest?
Why were length of stay data presented as means alone, and not
median? As length of stay is non-parametric the median is more appropriate
to evaluate what is happening to the group as a whole, whereas the mean
will only really inform on how the intervention reduces the outliers,
which I accept is also important, but is not the complete answer.
Why were no data presented on the incidence of complications and
death? Length of stay is being reduced all the time across most major
surgical groups through enhanced recovery strategies. Enhanced tissue
perfusion through correct fluid therapy should improve all of these
parameters. The lengths of stay prior to the intervention (in the selected
controls) seemed to be particularly lengthy indeed, and I note one centre
did not even show an improvement. In our hospital the median length of
stay in a published study of a selected high-risk major surgery was 8.5
I would reiterate that I believe firmly in the strategy of improving
surgical outcomes through ensuring optimal tissue perfusion. Using the
oesophageal Doppler is one way of doing this, and I would support it's
use. However, I don't believe this paper represents either true audit, or
properly conducted research, and am somewhat concerned about the massive
potential for bias in the results. This will not be of concern to the
company selling the Doppler device, and in their shoes I wouldn't be shy
about using this data. After all on their home page they say that "NICE
have stated the the CardioQ-ODM should be adopted for use in patients
undergoing major or high-risk surgery". Which is not quite the same as
the actual NICE recommendation that "the CardioQ-ODM should be considered
for use in patients undergoing major or high-risk surgery..", is it?
1. Wilson J, Woods I, Fawcett J et al. Reducing the risks of major
elective surgery. BMJ 1999; 318:1099-1103.
2. Davies S, Yates D, Wilson RJT. Dopexamine has no additional
benefit in high-risk patients receiving goal-directed fluid therapy
undergoing major abdominal surgery. Anesth Analg 2011; 112:130-8.
3. http://www.deltexmedical.com, 4th June 2011.
4. National Institute for Health and Excellence. CardioQ-ODM
oesophageal Doppler monitor. NICE medical technologies guidance 3. March
Competing interests: I have received honoraria for lecturing from Deltex Ltd (manufacturers of the oesophageal Doppler device) LiDCO Ltd (manufacturers of competing technology), and Fresenius-Kabi GmBH (manufacturers of intravenous fluid).
The easiest and cheapest adjunct to conventional CVP and arterial
pressure haemodynamic monitoring, is monitoring of central venous oxygen
saturation (ScvpO2). This can be monitored either intermittently with
central venous blood gas sampling, or continuously with oxymetric
catheters. An ScvpO2 of 75% indicates an adequate cardiac output. It is
true that the Oesophageal Doppler (ODM) provides additional information in
terms of corrected Flow time (FTc) to guide volume resuscitation, but in
truth changes in stroke volume can be inferred from SvO2 just as easily.
Oesophageal Doppler monitoring is subject to a number of important
limitations: in patients with proximal aortic disease or tortuosity it can
be inconsistent and unreliable; it is seldom helpful in the context of
atrial fibrillation. In patients where CVP monitoring is routine and may
be important for the administration of vasoactive drugs, ScvpO2 is a
fraction of the cost and provides most of the same information - enough
maybe to provide similar improvements to intraoperative care, and unlike
ODM suitable for use in awake post operative patients.
Competing interests: No competing interests