• Relative adrenal insufficiency [ Time Frame: 1 week ]
  

• Therapeutic failure [ Time Frame: 45 days ]
  
• Survival [ Time Frame: 45 days ]
  
• Variations in portal hypertension. [ Time Frame: 7 days ]
  
• Need for vasopressive drugs [ Time Frame: 45 days ]
  

Observational, prospective, open-label, in-patient study, that includes patients with upper gastrointestinal bleeding of variceal or peptic origin, and in patients with severe acute pancreatitis.

The adrenal function of every patient included will be evaluated in the first 24 hours of admission This assessment shall be performed using the corticotropin-stimulation short test (synacthen test), that includes serum and saliva determination of cortisol, in basal conditions and 30 and 60 minutes after the administration of 250 ug of corticotropin synthetic (Synacthen, Novartis Pharma AG, Basel, Switzerland).

The cortisol levels will be determined by competitive immunoassay using direct chemoluminescence technology (Bayer Corporation, Pittsburgh, PA, USA).

In patients with severe acute pancreatitis all of these determinations will be repeated at the third day of admission.

Several other clinical and biochemical features will be recorded.

Patients admitted to our centre that meet inclusion criteria and do not meet exclusion criteria

Inclusion Criteria (one of the following):

• Presence of acute upper gastrointestinal bleeding (variceal or peptic) presence of hypovolemic shock (defined as a systolic blood pressure <100 mmHg coupled with a heart rate> 100 ppm), or Hb < 80 g / L;
• Acute pancreatitis (diagnosed by clinical, radiological and biochemical features)with severity criteria (At least one of the following: index of Ranson > 3, APACHE II > 8 or CPR> 120 mg/L, Balthazar CT grade E)

Exclusion Criteria:

• Age <18 years and >80 years.
• Pregnancy.
• Patient refusal to participate in the study.
• Prior corticosteroids treatment(oral or topical).
• Treatment during the 30 days prior to inclusion with any of the following drugs: contraception, etomidate, ketoconazole, rifampin or phenytoin.
• History of cranial trauma or surgery.
• Any malignancy in treatment or progression.
• HIV infection.
• Prior known adrenal pathology.
• Patients not eligible to any active treatment because the existence of any clinical condition considered terminal.
• Patients with associated traumatic blood loss, or any other extraintestinal source of active bleeding.
• Burns.
• Patients who have been previously included in this study.