Bewertung von Risikofaktoren im Hinblick auf die Mortalität bei interventionsbedürftigen oberen gastrointestinalen Blutungen unter besonderer Berücksichtigung der chirurgischen Therapie

Evaluation of risk factors regarding mortality for upper gastrointestinal bleeding that requires intervention with special focus on surgery

by Eva Maria Weiß
Doctoral thesis
Date of Examination:2018-06-18
Date of issue:2018-07-17
Advisor:PD Dr. Alexander W. Beham
Referee:Prof. Dr. Dirk Raddatz
Referee:PD Dr. Thilo Sprenger
Persistent Address: http://dx.doi.org/10.53846/goediss-6927

 

 

Files in this item

Name:Promotion Eva Maria Weiß.pdf
Size:1.56Mb
Format:PDF

The following license files are associated with this item:

Abstract

English

Upper gastrointestinal Bleeding (UGIB) remains one of the most common emergencies with an overall mortality-rate of up to 30 %. These days most bleedings can be controlled by endoscopic intervention, whereas certain cases remain that have to be taken to surgery. These are cases of recurrent bleeding that cannot be controlled endoscopically, cases in which the source of bleeding cannot be identified during endoscopy or cases in which primary hemostasis can be achieved, but which have a great risk of rebleeding. Since clinicians lack reliable parameters based on which they can decide on the right therapy, 283 consecutive cases of UGIB, that required intervention, were investigated to identify potential risk factors for lethal outcome. Patients were grouped and defined as “overall”, “operated” and “non-operated”, “operated and dead” and “non-operated and dead” where appropriate and compared with each other. In some instances, variceal bleeding investigated separately due to a different therapeutical approach. Data collection included demographic and anamnestic data, laboratory results, endoscopic finings and methods of intervention as well as methods of surgical intervention and outcome. Finally Blatchford-, as well as clinical and complete Rockall-Score were calculated for each patient and compared for the accuracy of their prediction. Results show, that patients presenting only with melena had a 2.3-fold elevated chance to survive. In general patients with non-variceal bleeding suffering from liver cirrhosis had a 3.6-fold increased mortality risk (mortality rate (M): 43 %) and in patients suffering from cancer it was increased to the 2.3-fold (M: 34 %). Endoscopically treated non-variceal bleeding had a 5.1-fold increased risk of mortality when presented with shock symptoms (M: 44 %) and suffering from liver cirrhosis again increased the mortality risk to the 5.6-fold (M: 50 %). Operated patients with non-variceal bleeding were 5.8 times more likely to die when anticoagulated at the inition of bleeding (M: 43 %). This went up to a 17-fold increase when treated with glucocorticoids (M: 80 %, but only 4 out of 5 patients). Mortality in variceal hemorrhage depended strongly on existing comorbidities but not on certain associated risk factors. Partial gastrectomy was carried out as Billroth I (M: 17 %), Roux-Y-reconstruction (M: 50 %) and Billroth II (M: 67 %). Blatchford score achieved an error rate of 0 %, clinical Rockall score of 5 % and complete Rockall score of < 1 %, however these results require further validation since the range of patients was preselected. In summary, that patients presenting with only melena have a better chance of survival. Preexisting liver cirrhosis or suffering from cancer significantly increases the risk to die of UGIB in general. Patients with liver cirrhosis or shock symptoms should be considered for early surgery whereas patients with anticoagulation or premedication of glucocorticoids should better be treated with non-surgical alternatives. If partial gastrectomy is indicated, Billroth I should be preferred. For initial risk assessment Blatchford and complete Rockall score appear equally suitable, but this result requires further validation.
Keywords: surgery; risk assessment; endoscopy; upper gastrointestinal bleeding
Schlagwörter: obere gastrointestinale Blutung; Chirurgie; Endoskopie; Risikobewertung
 
Statistik