Point of Care Ultrasound Unhelpful in Undifferentiated Hypotension? The SHOC-ED trial & the 'Tale of the Eighth Mare'
Jon-Emile S. Kenny MD [@heart_lung] with illustrations by Carla M Canepa MD [@_carlemd_]
“History does not repeat itself, but it often rhymes.”
An 89 year old man with a 100 pack-year smoking history is admitted with weakness and inability to take anything by mouth. He was discharged 2 months prior after a treatment course for C. difficile colitis. He returns to the ED with confusion. On examination his eyes are sunken and has poor capillary refill. He is very thin, with a ‘pink-puffer’ phenotype and a ‘systolic murmur.’ His blood pressure is lower than his normal grade II hypertension and his blood work reveals acute kidney injury. The admitting medicine resident evaluates the patient’s IVC and finds it to be 1.5 cm in diameter with minimal respiratory variation – with this data, he withholds IV fluids and calls his attending.
Point-of-care ultrasound [PoCUS] is a rapidly expanding diagnostic and therapeutic modality in emergency departments [EDs] and intensive care units [ICUs] alike. Indeed, various protocols have been developed using PoCUS and there is early data that these protocols can change management. Nevertheless, protocol-based resuscitation has an imperfect past and data that PoCUS-based evaluation leads to patient-centered improvement is lacking on a large scale.
With this, the SHOC-ED trial was recently published in the Annals of Emergency Medicine.
What They Did
Adult patients were screened in the ED for persistent systolic blood pressure less than 100 mmHg or a shock index greater than 1.0; patients were excluded if a known cause of hypotension was apparent. Pregnancy, trauma and/or advanced life support were also exclusion criteria.
Those included were then randomized to either usual care without PoCUS or to receive usual care plus PoCUS. The patients were enrolled from emergency departments with clinicians well-trained in PoCUS and the algorithms followed were based upon the Abdominal and Cardiothoracic Evaluation with Sonography [ACES] and the Rapid Ultrasound for Shock and Hypotension [RUSH] approaches. In both groups patients were re-evaluated 60 minutes following the initial assessment. In total, 138 patients were randomized to PoCUS and 135 to control.
What They Found
About one-half of patients were found to have ‘non-cardiogenic,’ ‘sepsis-related’ hypotension. 76.5% of patients in the PoCUS group and 76.1% in the usual-care group survived to discharge. Both groups of patients received identical volumes of resuscitation fluid in the ED [1.6 L]. Further end-points such as inotrope use, CT scan acquisition, hospital admission and length of stay were all completely equivalent between groups.
In their classic 2008 paper describing the poor relationship between the central venous pressure [CVP] and both volume status and responsiveness, Dr. Marik and colleagues relayed the tale of the ‘Seven Mares.’ They found only one rigorous study revealing relationship between volume status and the CVP – in seven, standing, awake mares undergoing graded hemorrhage. With this very poor evidence base Dr. Marik, deservedly, decried the near ubiquitous use of the CVP for fluid management. From my perspective, the most interesting part of his analysis was the historical narrative he provided concerning the adoption of the CVP. He relayed the following from Forrester and colleagues – pioneers in hemodynamic monitoring – written in 1971:
“CVP monitoring in acute myocardial infarction is at best of limited value and at worst seriously misleading …”
Yet a small descriptive study from 1959 connecting a fall in CVP to acute blood loss provided ground-work for the explosion in CVP monitoring. Clinicians collectively blinded themselves to the complicated determinants of the CVP – only one of which is a patient’s ‘volume status’ – and opted to insert the CVP into simplified heuristics, ultimately spawning a wealth of unsophisticated, threshold-based ‘goal-directed’ therapies.
Poetically, as Mark Twain reminds us, our past echoes not in replica but rather in rhyme and the sweet song of ultrasonographic hemodynamic assessment found its earliest melodies roughly 30 years ago. At this time, Kircher and colleagues noticed a satisfactory relationship between CVP and inspiratory IVC collapse in healthy volunteers; as well, Cheriex and colleagues related the IVC size to dry weight in hemodialysis patients. Like the aforementioned 1959 investigation relating blood loss to right atrial pressure, these descriptive studies grossly oversimplified the complicated determinants of IVC size – yet these data set the framework for various ultrasonographic approaches to hemodynamic assessment.
Can we know that a patient needs or will respond to fluids with a small left ventricular cavity? No. Can we know that a dilated right ventricle excludes hypovolemia as a cause of shock? No. Is it possible that abnormal left ventricular function can benefit from intravenous fluids rather than inotropes? Yes. Can we know that a patient needs or will respond to fluids with a small IVC? No & No. Can a patient with a dilated or unvarying IVC respond to preload? Yes.
Yet the protocols studied in the SHOC-ED may lead the clinician to the contrary of the above. While it is possible that the standard-care [no ultrasound] arm of SHOC-ED contained clinicians able to determine hemodynamic state with superior clinical skills, this seems unlikely given how poor we are at predicting hemodynamics in this manner. In my view, it seems much more likely that the algorithms studied in SHOC-ED are inherently flawed and this is because all algorithms and thresholds ignore the complexities of individual hemodynamics couched within equally intricate and idiosyncratic respiratory mechanics [see figure 1]. Sadly, we have already heard these lyrics from the swansongs of the pulmonary artery catheter and the CVP. If trials using ultrasound insist on rhyming with the protocols of the past, we will find ourselves solemnly singing a eulogy for PoCUS – ‘The Serenade of the Eighth Mare.’
Figure 1: Many variables affect what you see on an echocardiogram. Many of these variables change rapidly in response to both disease and intervention(s) performed in the ED and ICU. These complexities do not lend themselves to facile, protocolized, interpretations in any one patient at any given time.
Return To Case
The patient is evaluated by the attending physician. She auscultates his heart and recognizes the murmur of tricuspid regurgitation. She explains to the resident the determinants of the CVP and how this relates to cardiac function and respiratory variation of great vein diameter. She also explains how this must be interpreted in light of clinical pre-test probability. The patient receives 500 mL of lactated ringers and his urine output and creatinine improve. Later that evening the patient has 4 large volume ‘horse barn odour’ Bristol type 7 bowel movements and is transitioned to gentle sodium bicarbonate boluses given his moderate non-anion gap metabolic acidosis.
Dr. Kenny is the cofounder and Chief Medical Officer of Flosonics Medical; he is also the creator and author of a free hemodynamic curriculum at heart-lung.org