GI Bleed? Occam vs Hickam

The other night a patient came in from a nursing home by EMS, reportedly unresponsive with respiratory with agonal breathing.  He had been intubated en route by paramedics.  Paperwork from the nursing home said Hep C, cirrhosis, and end stage metastatic liver cancer.  There appeared to be coffee ground emesis on his face and linens, and when asked, paramedics stated that he vomited on their initial intubation attempt.  He was hypotensive and unresponsive (HR was 90s, which was surprising since he wasn't on any rate reducing medications).  His rectal temp was 34.9.  An NG tube was placed which drained 500cc of dark blood/coffee ground stomach contents.  This was a GI bleed, and all the appropriate steps were initiated. Things were somewhat delayed as he had no vascular access (EMS had placed an IO that wasn't flowing), and we were unable to get a pulse ox or cardiac monitor to read on the patient even after switching out monitors several times.  Central access was obtained, and he received 2L NS rapidly, and 2 units of unmatched blood were ordered.  The blood and additional fluids were given through the Ranger.  GI and MICU were contacted.  A CC lytes (point of care VBG) returned with a pH <6.91, lactate >139, K 8.3, and, surprisingly, a HCT of 38.

This is where the the thought process changed.  This patient had a GI bleed, but is that what was causing his hypotension, altered mental status and hypothermia?  Our job as emergency medicine physicians is not to observe the way that something walks, and the way it quacks, then to say it's a duck.  Our job is to say this may be a duck, or a goose, or a swan, etc... and in critical patients, to provide both diagnostic and therapeutic interventions at the same time.  

Now this normal Hct could have been due to hemoconcentration or lag in a Hct that has yet to drop.  However this patient has the right to have as many medical problems as he damn well pleases.  Maybe the bleeding resulted from the acidosis and hypothermia from an already established sepsis.  Liver patients are obviously prone to sepsis from multiple sources, and his CXR showed bilateral opacities concerning for aspiration vs pneumonia.  The patient was also on Dexamethasone for his bone mets - is he adrenally suppressed?  Myxedema coma is also on the differential.  Was the hypothermia environmental?  The nursing home reported that his symptoms had only started 15 minutes prior to arrival, but maybe that is just when they noticed it.  He could have been uncovered, cold, and slowly deteriorating over the course of the day.  

In reality it was probably multifactoral.  This guy was really sick, and his family decided to withdraw care while he was in the ED.  It was a good case to remind me not to anchor on the first and "most obvious" diagnosis.