O44 new GI bleed nursing model & unit changes outcomes

Mills, Matilda, Glover, Ben, Abosamra, Mahmoud, Warren, Patrick, Ahmadi, Sara, Lambert, Gillian, Kinnaird, Nanette, Wright, Mark, Boger, Philip and Stammers, Matt (2022) O44 new GI bleed nursing model & unit changes outcomes. Gut, 71, [A26]. (doi:10.1136/gutjnl-2022-BSG.44).

Abstract

Introduction: a dedicated GI bleed and liver unit (GIBLU) was created within the GI inpatient footprint. Each bay is staffed by 1 specially trained nurse to 4 patients with continuous monitoring available, a readily accessible medical team and 24-hour access to senior gastroenterology support. 
Aim: to assess the effectiveness of this change across the hospital.

Methods: the pre-implementation cohort contained all patients undergoing inpatient therapeutic endoscopy between April 2019 &amp Feb 2020. The post-implementation cohort contained all cases between April 2021 – Feb 2022. GI bleeds outside the GIBLU were also analysed. Measured covariates included age, sex, Charlson comorbidity index (CCI) and hospital frailty risk score (HFRS). Outcomes of interest included 30-day mortality (30DM), length of stay (LOS), readmission and rescoped rebleed rates. Chi2, Mann-Whitney U and Logistic Regression (LR) were calculated using p<0.05.

Results: there were 225 scoped bleeds in the 2019 cohort and 359 in the 2021 cohort. There were no significant differences in age (p=0.8450) or sex (p=0.2057) between the 2019 and 2021 cohorts. Overall 30DM in 2019 was 10.6%(95%CI:6.6–14.7) vs 11.1%(95%CI:7.8–14.4) in 2021, difference 0.5%(95%CI: -0.3–1.3), (p=0.8565). A comparison of the GIBLU and non-GIBLU cohorts in 2021 is given in table 1: [O44 Table 1 Key Outcomes (means) not included].
To help control for confounding 30DM LR prediction models were constructed: Coefficients = GIBLU: -0.67 (p=0.1378), Age: 0.02 (p=0.2801), Male Sex: 0.89 (p=0.0857), CCI: 0.21 (p=0.1004), HFRS: -0.07 (p=0.7571). Although none of the covariates reached significance GIBLU is the only protective factor with any strength.

Conclusions: despite increased demands on the service, length of stay and 30DM were both significantly lower in the GIBLU cohort which cannot be explained purely by demographic differences between the cohorts. Expanding the scope of GIBLU locally in 2022 seems highly likely given these results.

This record has no associated files available for download.

Contributors

Download statistics