Aims
Distinguishing bleeding originating from the small versus the large bowel can be challenging due to overlapping clinical features, while hemostasis will be required in a subset of patients. This study aimed to identify factors associated with small bowel bleeding and predictors of the need for hemostasis.
Methods
A retrospective analysis was conducted on patients with confirmed lower gastrointestinal bleeding. Associations between clinical and laboratory parameters with bleeding location and the need for hemostasis were assessed using statistical tests and logistic regression analysis. The crude analyses revealed several associations. To account for potential confounding, we performed an adjusted analysis featuring all of the variables fulfilling the lenient statistical threshold of p= 0.01. The truly important associations are those that “survived” the adjusted analysis. Adjusted associations were considered significant using the conventional statistical threshold of p= 0.05.
Results
A total of 419 patients (242 males, 57.8%) with a median age of 76 years (IQR 65–83) were included. The bleeding source was located in the colon in 85.9% of cases, in the small bowel in 10%, and remained unidentified in 4.1%. Hemostasis was required in 13.8% of patients. The presence of melena and maroon-colored stools was strongly associated with small bowel bleeding (OR 18.5 and 7.0, respectively; p<0.001). A history of cardiovascular disease reduced the likelihood of small bowel bleeding by 87% (p<0.001). Higher hemoglobin levels were associated with a lower risk (15% reduction per 1 g/dL increase; p=0.027), while higher urea levels increased the risk (0.9% per mg/dL; p=0.01). Polypectomy markedly increased the likelihood of requiring hemostasis (OR ~36; p<0.001). Conversely, higher albumin levels significantly reduced this risk (53.4% per 1 g/dL increase; p<0.001), whereas anticoagulant use more than doubled it (p=0.049).
|
Small vs large bowel bleeding |
Sig. |
Exp(B) |
95% C.I.for EXP(B) |
|
|
Lower |
Upper |
|||
|
Referral |
,110 |
2,033 |
,852 |
4,849 |
|
Hematochezia |
,000 omnibus p |
|
|
|
|
Melena |
,000 |
18,537 |
5,651 |
60,803 |
|
Maroon coloured stool |
,000 |
7,107 |
2,881 |
17,532 |
|
Rectal bleeding |
,997 |
NA |
||
|
Bloody diarrhea |
,999 |
NA |
||
|
Cardiovascular disease |
,000 |
,127 |
,049 |
,331 |
|
Prior lower GI bleeding |
,144 |
2,115 |
,775 |
5,773 |
|
Platelets (K/μl) |
,917 |
1,000 |
,996 |
1,005 |
|
Albumin (g/dl) |
,499 |
,829 |
,481 |
1,428 |
|
Urea (mg/dl) |
,010 |
1,009 |
1,002 |
1,016 |
|
Hemoglobin(g/dl) |
,027 |
,846 |
,729 |
,981 |
Conclusions
Specific clinical and laboratory parameters are associated with both the bleeding site and the need for hemostasis. Their integration into clinical assessment may facilitate early recognition and targeted management of high-risk patients.