Predictive Model for Length of Hospital Stay of Patients Surviving Surgery for Perforated Peptic Ulcer

Background/Purpose: Many studies have assessed the major risk factors for mortality or morbidity in surgical patients with perforated peptic ulcer (PPU). The aim of our study was to focus on survivors and to investigate the predictive factors for length of hospital stay (LOHS) for surgical patients. Methods: The medical records of 195 patients who survived surgery were reviewed retrospectively. The following factors were analyzed: patient profile, associated medical illnesses, diagnostic methods, fever, preoperative shock, clinical data from the emergency room, delays in surgery, sites of perforation, operative methods, positive ascites culture and species of microorganism, postoperative infection, and non-infective postoperative complications. Results: Univariate analysis showed that the following factors were significantly related to longer LOHS: age > 65 years, liver cirrhosis, diabetes mellitus, history of peptic ulcer disease, number of comorbid diseases, preoperative shock, creatinine > 1.5 mg/dL, surgical delay of > 12 hours, and all postoperative factors except species of microorganisms in ascites culture. In a multivariate linear regression model, catheter infection, pneumonia, urinary tract infection, wound infection, bacteremia, non-infective abdominal complications, other infections, surgical delay of > 12 hours, and comorbidity were major factors associated with longer LOHS. Conclusion: A predictive model was established with nine factors that explained 71.6% of the variation in LOHS of patients who survived surgery for PPU. Several corrective methods based on the model can be devised by attending physicians to shorten LOHS.