RESUMO
BACKGROUND: Infection is a common cause of mortality within intensive care units (ICUs). Antibiotic resistance patterns and culture data are used to create antibiograms. Knowledge of antibiograms facilitates guiding empiric therapies and reduces mortality. Most major hospitals utilize data collection to create hospital-wide antibiograms. Previous studies have shown significant differences in susceptibility patterns between hospital wards and ICUs. We hypothesize that institutional or combined ICU antibiograms are inadequate to account for differences in susceptibility for patients in individual ICUs. METHODS: Culture and susceptibility data were reviewed over a 1-year period for 13 bacteria in the following ICUs: Surgical/Trauma, Medical, Neuroscience, Burn, and Emergency department. Antibiotic management decisions are made by individual teams. RESULTS: Nine species had sufficient data for inclusion into an All-ICU antibiogram. E coli and S aureus were the most common isolates. Seven species had significant differences in susceptibility patterns between ICUs. E cloacae showed higher rates of resistance to multiple antibiotics in the STICU than other ICUs. P aeruginosa susceptibility rates in the NSICU and BICU were 88% and 92%, respectively, compared to 60% and 55% in the STICU and MICU. Cephalosporins and Aztreonam had reduced efficacy against E coli in the NSICU, however remain effective in other ICUs. CONCLUSIONS: The results of this study show that different ICUs do have variability in antibiotic susceptibility patterns within a single hospital. While this only represents a single institution, it shows that the use of hospital-wide antibiograms is inadequate for creating empiric antibiotic protocols within individual ICUs.
RESUMO
BACKGROUND: Roughly 5% of patients with blunt abdominal trauma (BAT) have a blunt bowel and/or mesenteric injury (BBMI). Determining the need for operative management in these patients can be challenging when hemodynamically stable. Single center studies have proposed scoring systems based on CT findings to guide management. Our study aimed to determine the predictability of abdominopelvic CT scan (CT A/P) findings in conjunction with clinical exam to determine the necessity of operative intervention for BBMI. METHODS: Patients presenting from 2017 to 2022 to the University of South Alabama Level 1 Trauma Center after motor vehicle collision were retrospectively reviewed. Patients with CT findings suggestive of BBMI were further analyzed, noting CT findings, Glasgow coma scale (GCS), shock index, abdominal exam, operative or nonoperative management, and intraoperative intervention. RESULTS: 1098 patients with BAT underwent CT A/P. 139 patients had ≥1 finding suggestive of BBMI. 38 patients underwent surgical exploration and 30 had surgically confirmed BBMI. 27 patients required intervention for BBMI. Univariate analysis indicated that pneumoperitoneum (p < 0.0001), active extravasation of contrast (p = 0.0001), hemoperitoneum without solid organ injury (SOI) (p < 0.0001), peritonitis (p < 0.0001), and mesenteric stranding(p < 0.05) were significantly associated with intervention. CONCLUSION: In total, 30 patients had surgically confirmed BBMI. Active extravasation, pneumoperitoneum, hemoperitoneum without SOI, mesenteric stranding, and peritonitis were significant indicators of BBMI requiring intervention. CT and clinical findings cannot reliably predict the need for surgical intervention without ≥1 of these findings. Initial nonoperative management with serial clinical exams should be strongly considered to reduce incidence of nontherapeutic laparotomies.