Clinical Vs Endoscopic correlation of Upper Gastrointestinal Bleed

After enumerating the etiologies of UGIB, peptic ulcer constitutes the most common cause, out of which duodenal ulcer constitutes 25%, gastric erosions 28%, gastric ulcer 4%, duodenitis3% and esophageal ulcer 1%. This data correlates well with the UK registry of UGIB (Audit committee) which was published in BMJ by Rockall TA et al. The Canadian registry also has similar observations. Malignancy accounted for 4% of cases and varices constituted 20% of non ulcer bleed. UGI scopy was the diagnostic modality of choice and done in all cases. 74% of UGI scopy were done within 24 hours. Endoscopic intervention were done as appropriate and sclerotherapy was the most preferred procedure than EVL for esophageal varices. This choice was based on local preference, expertise and simplicity. Adrenaline injection was the preferred endotherapy for ulcer bleed. Compared data on the choice, EVL is the most common procedure worldwide and widely accepted for esophageal variceal bleed. Head to head trials comparing EST and EVL showed superiority of EVL over EST but visibility of the endoscopist after a torrential bleed is of concern during EVL (Laine L, Cook D et al). Total number of death is 9 (9%). We have formulated diagnostic criteria for major bleed as follows: 1. Death of the patient, 2. Who had transfusion requirement of more than 2 units, 3. Who had ICU stay more than 1 day, 4. Hospital stay more than 3 days. We have chosen a few patient parameters as potential risk factors for early prediction of major UGIB. The parameters considered are age, BP, HR, Hb at presentation,creatinine level, timing of endoscopy, endoscopic diagnosis and endotherapy. Using statistical analysis the following were observed : Age was considered as an important predictor of death. Age > 60 is associated with increased mortality (p<0.001). Age is not a predictor of increased ICU stay (P < 0.102), increased transfusion requirement (P<0.0305) or hospital stay (p<0.02). This is compared with Rockall et al which has similar mortality in > 60 years age group. Mortality increased with advancement of age (Blatchford et al). Blood pressure at presentation is found to have important predictor of death and major bleed. BP < 100/70 is associated with increased death (P <0.001), increased hospital stay (P < 0.001) and increased transfusion requirement (P<0.001). Similar way cold peripheries (Shock) at presentation predicts mortality, increased ICU stay, hospital stay and transfusion requirement (P<0.001). Study of Longstreth GF 1997 indicated that outcome of patient who presented with shock had higher mortality and constituted about 17 % of all bleeds. Heart rate >100 correlated well with mortality (P<0.001), ICU stay (P<0.001) and hospital stay (P<0.001). Hb < 8gms/dl predicted the need for increased transfusion (P<0.001), ICU stay (P=0), hospital stay (P=0) and death (P=0) Creatinine >2mg/dl correlated well with death, ICU stay, transfusion and hospital stay (P =0) Presence of coronary artery disease is surprisingly neither a predictor factor of mortality (P<0.171), ICU stay (P<0.954), transfusion (P<0.799) nor hospital stay (P<0.481) Presence of comorbidities like HT, CRF, arthritis, alcoholism or smoking did not predict death of major bleed. This is in contrast to study by Rockall which predicted high mortality for patients with comorbid illness. The discrepancy occurred because of skewed patients population towards younger age and low incidence of comorbid illness in our study. Timing of endoscopy and its possible improved outcome is a major controversy in management of UGIB. Various studies have produced variable results. In general UGI scopy performed <24hrs did not have any added advantage in reducing mortality or morbidity (Splejel BM et al ). Optimal timing of UGI scopy is put variably by different authors. But the consensus statement put it at < 24 hrs (Stering committee Rockall et al). Our series had 95 % of UGI done within 24 hrs. 78 % of patients who died had UGI done <24 hrs. 22 % had UGI done after 24hrs probably due to continued hemodynamic instability and resuscitation. The mortality in these subgroup is statistically significant to early UGI scopy group (P<0.004). The timing of UGI correlated well with ICU stay and hospital stay has expected. Patient who underwent endoscopy < 24hrs had good recovery despite major bleed (P<0.004) Endoscopic diagnosis of etiology emerged as the strongest predictor of mortality. Diagnosis of esophageal varices, bleeding peptic ulcer contributed 98 % of all mortality signifying the importance (P < 0.001). Endotherapy with sclerosant, adrenalin conveyed mixed results when correlated with mortality. Patient needing endotherapy were very sick contributing to overall mortality (P<0.001) but 69 % (16/23) of patient were salvaged by endoscopy (P<0.001). CONCLUSION : 1. 30% of the patients with UGIB had high mortality and morbidity. (Major UGIB). 2. Strong predictors of major UGIB are Age>60, HR>100, BP 2mgs/dl and shock at presentation. 3. Endoscopy is the most valuable investigation for diagnosis. UGI scopy done < 24 hrs predicted well with improved outcome. 4. Endoscopic diagnosis like esophageal varices, bleeding peptic ulcer carried high mortality. 5. Endoscopic therapy (injection for ulcer bleed, sclerotherapy for varices) salvaged 69 % of major bleed and improved prognosis.