P1528 - Implication of Blood Transfusion in Mortality in Non-Variceal and Variceal Upper GI Bleeding – Is It the Same for Both?

Monday, October 25, 2021

10:30 AM – 4:15 PM PT

Location: Shoreline Exhibit Hall

Has Audio

Award: Outstanding Research Award in the GI Bleeding Category (Trainee)
Award: Presidential Poster Award

Presenting Author(s)

Dhruvil Radadiya, MD

University of Kansas Medical Center
Kansas city, KS, United States

Award: Outstanding Research Award in the GI Bleeding Category (Trainee)

Award: Presidential Poster Award

Dhruvil Radadiya, MD¹, Viraj Shah, MD², Kalpit Devani, MD³, Don Rockey, MD⁴
¹University of Kansas Medical Center, Kansas city, KS; ²Florida Atlantic University, Boca Raton, FL; ³University of South Carolina, Greenville, SC; ⁴Medical University of South Carolina, Charleston, SC

Introduction: Volume resuscitation is the standard of care in all GI bleeding cases. Blood transfusion is one of the means to achieve the aforementioned objective. Blood transfusion often has been a surrogate of the severity and mortality in operative and traumatic settings. Out of all UGIB patients, variceal bleeders remain prone to further worsening of bleeding secondary to excess blood transfusion. Therefore, we hypothesized variceal bleeders have higher mortality than non-variceal bleeders secondary to blood transfusion.

Methods: We used Nationwide Inpatient Sample from the year 1998 to 2018. We identified all patients with UGIB using using ICD-9 and ICD-10 codes. We categorized these UGIB patients into two subgroups: Variceal Bleeding (VUGIB) and Non-variceal bleeding (NVUGIB). Our primary outcome of interest was in-patient mortality in patients receiving blood transfusions. Using logistic regression models we looked at the effect of blood transfusion on inpatient mortality in variceal and non-variceal bleeders separately. To study the effect of blood transfusion on inpatient death based on UGIB type (variceal vs. non-variceal), we only took patients who received a blood transfusion and did logistic regression after controlling for multiple covariates.

Results: Overall we identified 10,228,524 UGIB admissions, out of which more than half were NVUGIB. 42% of admission with UGIB were transfused during hospitalization and blood transfusion was more common in variceal bleeders (47% vs. 42%; p < .0001). Overall inpatient mortality for UGI bleeders was 4.2%. In patients with NVUGIB, blood transfusion was associated with 6% lower mortality (OR: 0.94; 95% CI: 0.92 - 0.97; p< .0001), while in patients with VUGIB blood transfusion was associated with 10% higher mortality (OR: 1.10; 95% CI: 1.05 - 1.16; p< .0001) as shown in Figure 1A & 1B respectively. Comparing variceal bleeders with non-variceal bleeders - after isolating admission receiving a blood transfusion - on multivariable logistic regression analysis 2-fold increase in mortality was noted amongst variceal bleeders (OR: 2.11; 95% CI: 2.03 - 2.21; p< .0001) as shown in table 2.

Discussion: Blood transfusion remains an independent marker of inpatient death in upper GI bleeding. However, patients with variceal bleeding after receiving transfusion had higher inpatient mortality than patients with non-variceal bleeding who received a blood transfusion. Thus, unnecessary blood transfusion for variceal bleeders should be discouraged.

Figure: Figure 1A. Factors associated with inpatient mortality in NVUGIB patients
Figure 1B. Factors associated with inpatient mortality in VUGIB patients

Characteristic	OR¹	95% CI¹	p-value
Age group
18 – 34	—	—
35 – 49	1.39	1.24 - 1.56	< 0.001
50 – 64	1.66	1.49 - 1.85	< 0.001
65 – 79	1.87	1.68 - 2.10	< 0.001
≥80	2.77	2.48 - 3.11	< 0.001
Female	0.85	0.83 - 0.88	< 0.001
Weekend admission	1.04	1.00 - 1.07	0.032
Elective admission	1.04	0.97 - 1.10	0.3
Race
White	—	—
Black	0.90	0.86 - 0.95	< 0.001
Hispanics	0.99	0.93 - 1.04	0.7
Other	0.98	0.93 - 1.04	0.6
Primary insurance
Medicare	—	—
Medicaid	1.18	1.12 - 1.25	< 0.001
Private	0.98	0.94 - 1.03	0.4
Self-pay	1.26	1.17 - 1.36	< 0.001
No charge	0.99	0.77 - 1.26	>0.9
Other	1.01	0.92 - 1.12	0.8
Income quartile
1	—	—
2	0.96	0.92 - 1.00	0.064
3	0.95	0.91 - 0.99	0.011
4	0.96	0.92 - 1.01	0.11
Region of hospital
Northeast	—	—
Midwest	0.71	0.66 - 0.75	< 0.001
South	0.82	0.77 - 0.86	< 0.001
West	0.88	0.83 - 0.94	< 0.001
Teaching hospital	0.92	0.88 - 0.96	< 0.001
Elixhauser comorbidity index	1.05	1.05 - 1.05	< 0.001
Hemorrhagic shock	2.34	2.24 - 2.44	< 0.001
Mechanical ventilation	14.6	14.1 - 15.1	< 0.001
EGD	0.39	0.38 - 0.40	< 0.001
VUGIB vs. NVUGIB	2.11	2.03 - 2.21	< 0.001
¹OR = Odds Ratio, CI = Confidence Interval

Table: Table 2. Factors associated with inpatient mortality in UGIB patients receiving blood transfusion

Disclosures:

Dhruvil Radadiya indicated no relevant financial relationships.

Viraj Shah indicated no relevant financial relationships.

Kalpit Devani indicated no relevant financial relationships.

Don Rockey indicated no relevant financial relationships.

Dhruvil Radadiya, MD¹, Viraj Shah, MD², Kalpit Devani, MD³, Don Rockey, MD⁴. P1528 - Implication of Blood Transfusion in Mortality in Non-Variceal and Variceal Upper GI Bleeding – Is It the Same for Both?, ACG 2021 Annual Scientific Meeting Abstracts. Las Vegas, Nevada: American College of Gastroenterology.