Surgical workload, risk factors and complications in patients on warfarin with gastrointestinal bleeding

This study aimed to assess surgical workload and risk factors for gastrointestinal bleeding in patients on warfarin admitted to a hospital.

Data was collected for all warfarinised patients admitted between April 2005 and October 2007 with gastrointestinal bleeding.

A total of 30 patients (average 80 years) were recorded. Indications for warfarin therapy were atrial fibrillation (80%), mechanical heart valve (6.67%) and embolic disease (13.33%). Fifty percent were admitted with an INR above therapeutic range and of these patients, 83% were on one or more medications known to potentiate the anti-coagulation effect of warfarin. Nine patients were also taking anti-platelet medication. Five of these nine had an admission INR within the intended therapeutic range. Thirteen patients received blood transfusions and had a significantly higher (p<0.05) INR (average 9) than the 17 patients not requiring transfusion (average 2.8). The average cost of transfusion per patient was pound470. None of the patients required acute surgical intervention. The average length of stay was 7 days, at a total cost of pound1444 per patient. Investigations found the cause of bleeding to be diverticulosis in 9 patients and neoplastic disease in 4 patients. Almost half of the patients received no investigation due to risks from co-morbidity.

Uncontrolled anti-coagulation, polypharmacy and age were overwhelming risk factors for major gastrointestinal bleeding. Our results show that adding anti-platelet therapy has to be clearly justified against the increased risk of bleeding. Cost to the surgical department was high and no patients required surgical or radiological intervention. WHAT IS ALREADY KNOWN ABOUT THIS TOPIC?: Warfarin is an important drug, but the complications of its use are difficult and expensive to deal with. Warfarin use is a risk factor for haemorrhage, and this commonly involves the gastrointestinal tract. The use of warfarin is set to increase as the population ages and atrial fibrillation and other cardiovascular risk factors become more prevalent. Consequently, one can expect a rise in warfarin-related gastrointestinal haemorrhage. WHAT DOES THIS ARTICLE ADD?: Our study aimed to assess the burden of gastrointestinal haemorrhage secondary to warfarin on our surgical department (which was high), and also to assess what the risk factors for haemorrhage for patients on warfarin. One of the risk factors we uncovered was polypharmacy, particularly involving anti-platelets e.g. aspirin. We highlight the need for further guidance with regards to managing patients on warfarin, and suggest possible solutions to the problems uncovered.