Goals: In this study we analysed the patients admitted with orthopaedic problems who had coffee ground vomitus for incidence, risk factors, investigations and the management. This project was to highlight this significant but rather neglected problem and to draw up local guidelines in the prevention.

Methods: This is a prospective study conducted for a period of six months from 1st of July 2005 to 31st of Dec 2005. All patients admitted with Orthopaedic problems who had coffee ground vomitus were included in the study after confirmation with haemoccult test. Their case notes were studied to identify the risk factors, preventive measures that could have been taken and their management.

Results: There were 34 (2.3%) patients who had coffee ground vomitus, out of 1427 orthopaedic admissions during the study period. There were 14 (41%) men and 20 (59%) women. The mean age was 73.7 yrs in men and 82.2 yrs in women. This problem was more common in 8th decade with 15 patients (45.5%). There were 6 patients each in 7th and 9th decade, 5 patients in 6th decade and only 2 patients in 5th decade.

There were 19 (55.5%) trauma admissions with fractures and 15 (45.5%) elective admissions. There were 12 (35.2%) patients with previous gastric problems. There were 20 (59%) patients who were on gastric irritant medications, out of which only 5 (25%) were on gastro protective medications. All 34(100%) patients were on low molecular weight heparin for thromboprophylaxis. There were 2 patients on steroids and 2 patients on warfarin.

Coffee ground vomitus occurred preoperatively in 4 (13.4%) and postoperatively in 26 (86.6%). It happened with in the first six hours after surgery in 25 (96.5%) patients. Only in one patient it happened after 3 weeks.

All patients were kept nil by mouth, started on fluid resuscitation and intravenous ranitidine followed by oral omeprazole. Patients who were haemodynamically unstable were investigated by endoscopy. 17 (50%) patients had oral gastroduodenoscopy. 2 patients had blood transfusion because of significant drop in haemoglobin and one died before the transfusion was started.

There were 5 (14.7%) deaths in our study group. The cause of 2 deaths was directly related to gastrointestinal bleeding and the other three were confirmed to have had concurrent chest infection.

Conclusions: Gastro intestinal bleeding is a neglected but not an uncommon problem in orthopaedic patients. Identification of high-risk patients and implementation of preventive measures could avoid this potential life threatening complication. We recommend withdrawal of gastric irritants and co-prescription of gastro protective medications for high-risk orthopaedic patients.