RESULTS
Data of 197 patients admitted to our Hospital between January 2017 and May 2018 with the diagnosis of AGI bleeding were retrospectively analyzed.
When demographic data of 197 patients followed because of AGI bleeding were analyzed, 118 were male and 79 were female. Mean age of the 197 included patients was 68. When bleeding sites of the patients followed with AGI bleeding were classified, 132 patients had upper GI bleeding, 57 had middle-lower GI bleeding and bleeding site could not be established in 8 patients because endoscopy could not be carried out due to of vital instability.
It was detected that chronic liver disease was present in 12.1% (n:24) of the patients. It was observed that acute renal insufficiency developed in 46 (23.4%) of the patients after GI bleeding. It was detected that 53 patients (26.9%) had prior GI bleeding history.
The mean Hb level of 197 patients at the time of presentation is 9.3 g/dL. When the patients who were exitus (n:28), transferred to external medical center intensive care unit (n:3) and who refused the treatment were subtracted from 197 patients, the mean Hb level of the remaining 161 patients at the moment of discharge was found 10.1 g/dL. Anemia according to WHO criteria during discharge was detected in 147 patients (91,3%). It was seen that anemia treatment was given to only 15 (10,2%) of the 147 patients in whom anemia was detected. Within the 3 months following discharge, only 65 (40,4%) of the 161 patients who had presented with A-GIS bleeding and were followed-up came for the control follow-up visit, the mean Hb level of these patients were found 10.7 g/dL and it was seen that anemia continued in 80% of the patients at present.
Demographic data, AGI bleeding sites, Hb levels and anemia status of 197 patients with AGI bleeding are summarized in Table 1.
Hundred and thirty-one patients underwent upper GI endoscopy, 3 enteroscopy, 45 lower GI endoscopy, 8 upper and lower GI endoscopy, 2 upper, middle and lower GI endoscopy, and 8 patients could not undergo any procedures because of vital instability. The three most frequent causes of upper AGI bleeding were duodenal ulcer, gastric ulcer and esophageal variceal bleeding, respectively, and causes of upper AGI bleeding are summarized in Table 2 in order of frequency. The three most frequent causes of lower-middle AGI bleeding were hemorrhoids, diverticulae and angiodysplasias, respectively, and causes of lower-middle AGI bleeding are summarized in Table 3 in order of frequency.
Sixty nine patients underwent an endoscopic therapeutic intervention (47.8% sclerotherapy, 15.9% argon plasma coagulation, 5.8% clip application, 5.8% band ligation, 24.6% mixed procedure). Mean hospital stay was 10±14 days, and 67% of patients underwent transfusion with 3.5±5.89 units of erythrocyte suspension during hospital stay. Of 197 patients followed with AGI bleeding, 28 died and 3 of 169 patients were referred to the intensive care unit of another hospital, 5 refused therapy, and 161 were discharged after their post-hospitalization therapies were completed. Mean Hb levels of patients after follow up of AGI bleeding was 10,1 g/dL.
147 patients (91.3%) (F 91.1%, M 91.4%, upper GI bleeding 92.8%, middle GI bleeding 100%, lower GI bleeding 87.2%) had anemia during discharge. When frequency of anemia was viewed in terms of gender, the frequency of anemia having been found as 85/93 (91.3%) in male patients and 62/68 (91.1%) in female patients, meaningful relationship was not found between prevalence of anemia and gender (p>0.05). Having been found as 103/111 (92.8%) in the patient group with upper AGI bleeding and as 44/50 (88%) in the patient group with lower-middle AGI bleeding, meaningful relationship was not found in respect of frequency of anemia prevalence according to bleeding site (p>0.05).
It has been seen that there is a meaningful relationship between high BUN at presentation and elongated hospitalization with frequency of anemia prevalence (p< 0.05).
It has been seen that there is a meaningful relationship between incidence of anemia and comorbid disease in the patients followed with GI bleeding (p<0.05) (Table-4). As this situation may be secondary to pre-existing chronic disease anemia, it suggests multiple drug use causing complicated lesions that will create bleeding susceptibility in GIS and chronic disease anemia and iron deficiency anemia could be overlapping. Reviewing the anemia parameters in the patients will be helpful for us to reveal this situation more clearly.
Fifteen patients noted to have anemia (10.2%) were prescribed oral (n=12, 7.8%) and parenteral (n=3, 1.9%) iron therapy during discharge. A mean increase of 2,96 gr/dL was observed in Hb levels of the patients given oral treatment in a three-month period. Only 1 patient of the 3 patients given IV treatment came to control visit and it was seen that there was 1 gr/dL increase in Hb level.
65 of 161 patients (40.3%) were re-evaluated in the first three months following discharge and anemia persisted in 80% of these. Of 15 patients receiving treatment for anemia, 11 came to control visits in the first 3 months after discharge and anemia was noted to resolve in 4 (36.3%) patients. Again, 54 of 132 patients not receiving treatment for anemia following discharge came to control visits in the first 3 months after discharge and anemia was noted to resolve in 9 (16.6%) patients. None of the patients followed after discharge required blood transfusion.