The vital threat of an upper gastrointestinal bleeding: Risk factor analysis of 121 consecutive patients

Table 1 Reason for the occurrence of upper gastrointestinal (UGI) bleeding

Reasonfor bleeding	Duodenalulcer	Gastriculcer	Erosivemucosal lesion	Esophagealvaricosis	Mallory-Weiss-syndrome	Cancer	Coagulopathy	Lymphoma	Benign tumor	Unknownreason
%	29.8	28.9	16.5	9.9	6.6	4.1	3.3	1.7	1.7	0.8
Number (n = 125)1	36	35	20	12	8	5	4	2	2	1

¹Four patients presented with both duodenal and gastric ulcers with each showing an endoscopic evidence of bleeding and thus each ulcer was counted as a whole.

Table 2 Logistic regressions analysis of covariables recorded at admission of 121 patients considering their ability in predicting the patients’ lethality

Covariable	P	Statistical inclusionin model	Odds ratio	95% confidenceinterval
Cirrhotic liver disease	0.005	Yes	4.5	1.4-13.8
Hb	0.256	No
Medication	0.419	No
PT	0.672	No
Age	0.822	No
Gender	0.879	No

Table 3 A logistic regression analysis evaluating 14 prognostic parameters in their ability to increase the predictability of lethality of 67 patients with UGI bleeding resulting from duodenal or gastric ulcer

Covariable	P	Statistical inclusion inmodel	Odds ratio	95% confidence interval
Cirrhotic liver disease	0.029	Yes	5.9	1.1-34.4
Gender	0.065	Yes	4.8	1.0-34.8
Duodenal ulcer (bulbus back wall)	0.135	Yes	3.5	0.6-20.7
Forrest 1 b	0.175	No
Forrest 2 b	0.272	No
Hb	0.285	No
PT	0.641	No
Medication	0.693	No
Forrest 3	0.721	No
Forrest 1 a	0.748	No
Forrest 2 a	0.782	No
Vascular stump	0.852	No
Age	0.926	No
Gastric ulcer (small curvature)	0.964	No

Table 4 Logistic regressions analysis indicating the predictive value of 12 covariables in predicting the lethality of an UGI bleeding

Covariable	P	Statisticalinclusionin model	Odds ratio	95 % confidenceinterval
Cirrhotic liver disease	0.005	Yes	5.7	1.8-19.3
Duodenal ulcer (bulbus back wall)	0.092	Yes	3.4	0.7-14.7
Bleeding recurrance	0.271	No
Hb	0.341	No
Medication	0.345	No
Esophageal varicosis	0.401	No
Number of operations	0.42	No
Substitution of RBCs	0.436	No
PT	0.625	No
Gastric ulcer (small curvature)	0.799	No
Gender	0.809	No
Age	0.863	No

Table 5 Subgroup analysis of 21 patients with UGI bleeding and liver disease (Evaluation of the influence of 9 covariables on the lethality within the group)

Covariable	Odds ratio	95% confidenceinterval	P
Hb	9.75	0.8-121.8	0.077
Bleeding recurrence	4.5	0.6-32.3	0.135
Number of operations	4.5	0.5-37.4	0.164
Substitution of RBCs	1	0.1-7.5	1
Gastric ulcer (small curvature)	1	0.1-13.4	1
PT	0.72	0.1-5.2	0.744
Gender	0.68	0.1-5.5	0.718
Age	0.3	0.03-3.3	0.322
Esophageal varicosis	0.17	0.02-1.8	0.137