Gastrointestinal Bleeding in the Critically Ill: Is there a Difference in Outcomes in Patients Presenting During the Weekday vs Weekend

1089

Purpose: Published literature supports differences in outcome of patients hospitalized on a weekend compared with weekdays. The purpose of this study was to evaluate any “weekend effect” in GI bleed patients admitted to the Cooper intensive care unit as to treatment modalities performed and outcomes.

Methods: A retrospective review of data for patients admitted to a medical surgical intensive care unit during 9/07-8/09. Patients were classified as weekday (Sunday 17:01-Friday 17:00) or weekend (Friday 17:01-Sunday 17:00) admission. Treatment performed while in the ED and GI consults were captured using hospital computer systems. Patients were evaluated for a variety of outcome parameters, including mortality and length of stay.

Results: A total of 102 patients (pts) were included: weekday admission n=68, weekend admission n=34. Age, gender, race, APACHE II score, chronic GI disease profile, advance directive, and life support status were similar in both groups. Pts admitted on weekends had longer ICU length of stay (LOS) than weekday patients (p=0.008). No significant differences between the two groups were observed for ancillary procedures (diagnostic/therapeutic upper/lower endoscopy) in ED, ancillary procedures in ICU, pRBCs transfused, hospital LOS, and mortality.

Conclusion: This study shows patients admitted during weekend with either upper or lower GI bleed had longer ICU LOS with no difference in hospital LOS or mortality. One possible explanation includes more severe bleeding in weekend admit patients despite similar severity scores, supported by a trend for greater percentage of therapeutic endoscopies (weekend upper endoscopy 20.6% vs. weekday 8.8%, weekend lower endoscopy 2.9% vs. weekday 1.5%). Another potential explanation for the longer ICU LOS is that throughput issues affect more timely discharges from the ICU over the weekend.

Table: [1089] Treatment modalities

Table: [1089] Outcomes