Evaluation of empirical antibiotic treatment adequacy in critically ill sepsis patients: impact on outcomes and their predictors / Khalid Ahmad Ali Al-Sunaidar

Inadequate antibiotic treatment has an impact on high mortality. The administration of adequate empirical antibiotics in the management of sepsis in intensive care units has been the cornerstone for treatment success. This study was to investigate the adequacy of empirical antibiotic in critically ill sepsis adult patients and the impact on outcomes such as mortality, severity index and ICU length of stay before being discharged. It had compared the outcomes of most common used antibiotics, supportive treatments with their predictors. It also examined the characteristics of patients admitted in ICU with sepsis. This is a retrospective observational cohort study and it was conducted in tertiary hospital Sungai Buloh -Selangor. The data have been obtained from computerized system /medical records of hospital for patients who were diagnosed as sepsis (based on the diagnosis or sepsis criteria) or have sepsis symptoms based on systemic inflammatory response criteria and being admitted from 2011-2015. Data was retrieved from computer systems, it was double screened and checked by the researcher with patient’s files in department of records. The adequacy was determined based on ICU guidelines, bacterial sensitivity patterns, dose, frequency, creatinine clearance and time of empirical antibiotics. APACHE II score was determined with online clinical calculator. Out of 228 ICU adult’s patients, 193 (84.6%) died with 119 (52.2%) male and 74 (32. 5%) female respectively. The mean ICU-length of stay (LOS) was 9.86 ±8.96 days, while the mean APACHE II score was 29.59±7.49 points. The inadequate empirical antibiotics (non-AEA) was significantly associated with mortality and ICU-LOS (P<0.005). In multivariable (MV) logistic regression, only the model of non-AEA was a predictor for non-survival OR=.395 (95% CI 0.184- 0.85) (P=0.004). In simple linear regression, the model of non-AEA was a predictor of ICU-LOS (R²=.055, 95% CI -7.184- -2.114). In MV linear regression, four variables were more likely to be associated with reduction of APACHE II scores, such as mild Glasgow coma scale, CNS source of infection, cefepime 2gm dose every 8 hours. and albumin received treatment respectively. While only two variables were more likely associated for increment of APACHE II scores septic shock diagnosis and continuous renal replacement therapy (CRRT) supported patients respectively (R²=0.779). In MV linear regression, six variables were more likely associated as predictors for the increment of ICU-LOS (Imipenem 250mg dose every 12 hours., Intermittent dialysis, Enterococcus faecalis bacterial infection, Acinetobacter infection with multiple resistance organisms (AC-MRO) infection, deep vein thrombosis (DVT) disease and surgery as source of infection) respectively. Besides, in MV cox regression, there was one variable associated with risk of mortality the dose of meropenem 1000mg/8hr/day [HR 19.254 CI95 % (3.124-118.647) (P=.001)]. Meanwhile, other four variables have a protective effect, such as dose of AB exceeds the recommended dose based on. CrCL. [HR .186 CI 95 % (.040-.868) (P=.032)], four organ dysfunctions [HR .128 CI 95%(.025-.654) (P=.014)], AC MRO bacteria [HR .102 CI 95%(.013-.780) (P=.028)] and intermittent dialysis [HR .027 CI 95% (.002-.321) (P= .004)] respectively. AEA was predictor for survival. The septic shock and CRRT were predictors to increase the APACHE II. Meanwhile, predictors for ICU-LOS were Imipenem dose 250mg/12hours, Intermittent dialysis, Enterococcus faecalis, Acinetobacter with MRO infection, DVT and surgery as source of infection. The implementation of antimicrobial stewardship programs, reduces the emergence of MDR infections and appropriate empirical antibiotics would improve the outcomes of sepsis ICU patients.