Determinants of postoperative length of hospital stay following emergency abdominal surgery in the gastrointestinal surgical unit at Mulago National Referral Hospital
Abstract
Introduction: Patients who have undergone emergency abdominal surgery tend to have prolonged postoperative hospital stay attributable to several factors within the perioperative period. In addition to indicating an inefficient health system, prolonged hospital stay denies other patients the access to similar health services and affects the hospital resource distribution. There is a paucity of data in Uganda, hence the need to identify areas in the perioperative period where intervention may be deployed to improve outcomes.
Objectives: To determine the average length, and identify the determinants of postoperative stay following emergency abdominal surgery
Methodology: An observational prospective study was conducted; 198 participants were recruited through consecutive sampling. Continuous variables and postoperative LOS were summarized into median and interquartile ranges. The LOS was categorized into short (1-3 days), medium (4-6 days) and long (more than 7 days) and proportions of each determined. Data analysis was done using STATA version 14.0.
Results: Having colorectal surgery OR 4.24 (95% CI; 1.26 to 14.26 P= 0.019); duration of surgery of more than 2 hours OR 6.39 (95% CI; 2.02 to 20.20 P < 0.002), presence of drain OR 6.37 (95% CI; 1.24 to 32.84 P= 0.027); mobilization out of bed at 2 days or more ([OR 4.18 (95% CI; 1.87 to 9.37 P = 0.001 or OR 19.59 (95% CI; 5.25 to 73.09 P < 0.001] respectively) as well as development of late complications OR 143.43 (95% CI; 42.43 to 487.69 P < 0.001) were found to be significant predictors for postoperative LOS.
Conclusion: The median postoperative length of stay following emergency abdominal surgery was 5 days (IQR 3-8 days). The postoperative length of stay was independently predictable by colorectal surgery, duration of surgery, presence of drains, mobilization out of bed, and late complications.