Mid Sweden University

OBJECTIVE: To evaluate the prevalence of impaired primary haemostasis in patients with acute upper gastrointestinal bleeding.

DESIGN: Prospective open study.

SETTING: County hospital, Sweden.

SUBJECTS: 160 consecutive patients admitted with haematemesis or melaena, or both.

INTERVENTIONS: Diagnosis verified on admission by endoscopy. Consumption of non-steroidal anti-inflammatory drugs (NSAID), transfusion requirements, presence of chronic liver disease, and age were recorded.

MAIN OUTCOME MEASURES: Skin bleeding time was measured on admission with a modified Ivy method using the Simplate II bleeding time device; von Willebrand factor and routine haematological tests were analysed.

RESULTS: In 31 of the patients (19%) the skin bleeding time was prolonged (over 10 minutes), and these patients also had a significantly lower packed cell volume on admission. In all patients the concentration of von Willebrand factor was increased which might explain why there was no significant difference in bleeding time between those who did and those who did not take NSAIDs.

CONCLUSION: We recommend measurement of bleeding time in patients who present with haematemesis or melaena, or both, to detect those with impaired primary haemostasis.