The assessment of fluid responsiveness in haemodynamically unstable patients.

My good friend Gavin Denton (@dentongavin) has once again surpassed himself in analysing some of the evidence for the passive leg raise. Over to you Gav (I think I love him….ssshhhh).

 

Background.

Clinical examination and findings have long been used to guide fluid resuscitation. Examination would be used to judge the patients’ fluid status. Tibby et al (1997) found that senior clinicians in critical care could only determine the fluid status of a given patient 50% of the time. Clinicians’ abilities to estimate cardiac index in ventilated children and infants. 1997. 

Hanson et al (2013) identified similar findings in haemodynamically unstable patients with malaria. The reliability of the physical examination to guide fluid therapy in adults with severe falciparum malaria: an observational study. 2013  

Since the 1970’s, static, chamber pressure based monitoring of the circulation has been used to guide fluid therapy, using surrogates to make assumptions of where the ventricles end diastolic volume sits on the Starling’s curve (Osman et al, 2007). Cardiac filling pressures are not appropriate to predict hemodynamic response to volume challenge. 2007.

Again, over the last ten years there is growing evidence that these indices, such as central venous pressure, and static pressures derived from pulmonary artery catheters, do not reflect the position of the cardiac ventricles on the Starling’s curve (Marik et al, 2008). Does Central Venous Pressure Predict Fluid Responsiveness?A Systematic Review of the Literature and the Tale of Seven Mares. 2008.

One large randomised control trial (Harvey et al, 2005) on the risk benefit of pulmonary artery catheters, has seen a large reduction in their use subsequently (Marik, 2013). Assessment of the clinical effectiveness of pulmonary artery catheters in management of patients in intensive care (PAC-Man): a randomised controlled trial. 2005Noninvasive cardiac output monitors: a state-of the-art review. 2013.

The passive leg raise (PLR) has long been used as an initial treatment for hypotension; the technique has the potential to auto-transfuse 200-300mls of blood from the lower extremities, and into the central circulation (Monnet et al, 2006).  Passive leg raising predicts fluid responsiveness in the critically ill. 2006.

Some studies have suggested that, when a PLR is combined with a flow or volume based mode of CO monitoring, the observed change in SV can be used to represent a reversible VE to assess fluid responsiveness (Boulain et al, 2002). Changes in BP induced by passive leg raising predict response to fluid loading in critically ill patients. 2002.

In addition, researchers argue that comparing pre and post PLR stroke volume not only predicts fluid responsiveness, but also allows monitoring methods such as PPV and SVV to be applied to patients outside of the operating theatre environment (Préau et al, 2010). Passive leg raising is predictive of fluid responsiveness in spontaneously breathing patients with severe sepsis or acute pancreatitis. 2010. This would include patients with arrhythmia and spontaneous breathing.

Following are two reviews of studies employing the passive leg raise, combined with a form of stroke volume monitoring to predict fluid responsiveness. The reviews use CASP diagnostic tool as the appraisal format.

Diagnostic Appraisal of Marik et al 2013

QuestionAnswerAdditional Comments
1. Was there a clear question for the study to address?Yes.Use of PLR in assessment of fluid responsiveness and utility of carotid Doppler.
2. Was there a comparison with an appropriate reference standard?Yes.Use of VE and monitoring change in stroke volume.
3. Did all patients get the diagnostic test and reference standard?Yes.All patients received VE following PLR.
4. Could the results of the test have been influenced by the results of the reference standard?NoPLR is a relatively immediately reversible test and would not effect the subsequent VE. However, no documented period of equilibration between reversal of PLR and reference test.
5. Is the disease status of the tested population clearly described?Can't tellNo table of demographics of the study population. Some data on age range, inotropic support and use of invasive ventilation provided.
6. Were the methods for performing the test described in sufficient detail?Yes.Study protocol stated, no data on adherence to the protocol.
7. What are the results?N=34.
PLR is sensitive and specific to fluid responsiveness compared to VE.
Sensitivity= 94%
specificity= 100%
Likelihood ratio += inf
Likelihood ratio -= 0.06
Positive predictive value= 100
Negative predictive value= 94

53% of patients FR.
PLR 94% and a specificity of 100% for predicting VR compared VE. No LR or PV provided. Data table presented allows calculation of 2x2 statistics. No p value applied to PLR vs VE in responders or none-responders. No provision of confidence intervals.

Statistical significance was only applied to comparisons of CO monitor rather than between PLR and the reference standard.
8. How sure are we about the results?Small study, no blinding, retrospective chart review data.No blinding of examiner, data collector or analyser to reference standard test. No data representing matching of responders vs non responders regarding demographics. Unclear number of examiners. No discussion of data extraction form
9. Can the results be appied to your patients/population of interest?Yes.Study population meets PICO question. Population a mix of both ventilator and non-ventilator dependent. No table presentation of demographics however.
10. Can the test be applied to your population of interest?Yes.PLR test logistically simple to perform, no cost implication. Bioreactance cardiac output monitor is a more recent technology, but stroke volume measurement universal to cardiac output monitoring.
11. Were all outcomes important to the individual or population considered?Yes.Assessment of fluid responsiveness central part of critical care organ support.
12. What would be the impact of using this test on your population?Potential for reduction in unnecessary fluid challenges and reduced fluid balance.May allow broader application of none invasive CO monitoring. Aid to decision making in patients with heart failure. Would be useful in application of goal directed therapy protocols.

Diagnostic Appraisal of Preau et al 2010

QuestionAnswerAdditional comments
1. Was there a clear question for the study to address?Very clear research question in the abstract, different questioned posed laterTwo questions:
- Does PRL predict FR?
- Are PP, SV (via cardiac echo) and femoral artery doppler equivalent in measuring FR.
2. Was there a comparison with an appropriate reference standard?VE was the reference standard for FR.Use of VE and monitoring change in SV.
3. Did all patients get the diagnostic test and reference standard?Yes.All patients received VE following PLR.
4. Could the results of the test have been influenced by the results of the reference standard?NoPLR is a relatively immediately reversible test and would not effect the subsequent VE. Period of equilibration between PLR and reference test part of the protocol.
5. Is the disease status of the tested population clearly described?YesDetailed demographics of diagnosis, sex, age and parameters relating to signs of hypoperfusion. All patients were none ventilated.
6. Were the methods for performing the test described in sufficient detail?Detailed protocol.Well described protocol of PLR and VE with prescribed period of equilibration between tests.
7. What are the results?n=34.
PLR= p value <0.001 for SV increase with same p value for VE. measured with transthoracic echo.
PLR SV for pulse pressure and femoral doppler also had p value <0.001 and correlated with sv increase by thoracic echo p <0.001.

41% of patients were fluid responders


PLR 10% sv change threshold.
Sensitivity= 86%
specificity= 90%
Likelihood ratio + = 8.6
Likelihood ratio - = 0.16
Positive predictive value= 86%
Negative predictive value= 90%
Overall PLR had a 86% sensitivity/90% specificity in predicting FR compared reference standard
The raw data is presented as mean/standard dilation so verification of results can not be independently calculated.
ROC graph is presented showing inflection points of FR with the various monitoring techniques.
No confidence intervals presented.
Some data presented for intra observer variability SV measurements.
8. How sure are we about the results?Retrospective chart review study.
Only demographic data presented in raw format.
P values presented, but no confidence intervals.
Number of observers not reported.
No blinding of results between PLR and VE.
No statement regarding completeness of data.
9. Can the results be applied to your patients/population of interest?YesNone ventilated patients, but still within the population of interest. All patients have sepsis or SIRS which again fits within the population of interest.
10. Can the test be applied to your population of interest?YesPLR test logistically simple to perform, no cost implication. Transthoracic echo not easily available in critical care, however, used as a reference technique, shows various haemodynamic modes of monitoring can detect response to PLR.
11. Were all outcomes important to the individual or population considered?YesAs perviously stated re-mode of haemodynamic monitoring.

Assessment of FR central part of critical care organ support.
12. What would be the impact of using this test on your population?Potential for reduction in unnecessary fluid challenges and reduced fluid balance.May allow broader application of none invasive CO monitoring. Aid to decision making in patients with heart failure. Would be useful in application of goal directed therapy protocols.

Pooled 2 x 2 contingency results Marik et al (2013) and Preau et al (2010) N =68

Sensitivity90.6%
Specificity94.4%
Positive predictive value93.55%
Negative predictive value91.89%
Positive likelihood ratio16.31
Negative likelihood ratio0.10
Pre-test probability47%

 

The bottom line.

  • Half of patients who have signs of haemodynamic compromise will not be fluid responsive.
  • Measuring the difference in stroke volume or cardiac before and after a passive leg raise is a highly sensitive and specific test of fluid responsiveness.
  • The mode of cardiac output monitoring probably does not matter as it is the percentage change that is important rather than a specific number.

 

Originally posted 2015-01-14 14:00:53. Republished by Blog Post Promoter

Share →

Leave a Reply

Your email address will not be published. Required fields are marked *