[Development of a predictive model for hospital mortality and re-admission in a cohort of infected patients that require hospitalization]. / Desarrollo de un modelo predictivo para mortalidad y otro para reingreso hospitalario en una cohorte de paciente con infección que requieren hospitalización.

OBJECTIVE: The aims of the study were: to develop a predictive model for hospital mortality and another for hospital re-admission, to identify the impact of antibiotic delay in the mortality rate and, to report the rate of inappropriate antibiotic therapy. METHODS: A cohort and retrospective study was conducted at the HM Sanchinarro University Hospital during the period September 1st, 2012 to March 31th, 2013. The inclusion criteria were: age> 18 years, hospital admission from the ED with a diagnosis of bacterial infection. The exclusion criteria were: suspected viral infection, negative bacteriological cultures, life expectancy less than 6 months, lack of clinical information, assistance exclusively by the trauma emergency department. Two logistic models were made (hospital mortality and hospital re-admission). RESULTS: A total of 517 patients were included. The final mortality model (30 deaths) include the following variables: respiratory rate (OR 1.12; IC95% 1.02; 1.22), oxygen saturation (OR 0.92; IC95% 0.87; 0.98), creatinine (OR 2.33; IC95% 1.62; 3.36), COPD (OR 3.02; IC95% 1.06; 8.21), cancer (OR 3.34; IC95% 1.07; 9.98) and chemotherapy in the last 3 months (OR 4.83; IC95% 1.54; 16.41). The final model for hospital re-admission (28 re-admissions) include the following variables: hepatopathy (OR 5.51; IC95% 1.57; 16.88), GPT (OR 1.005; IC95% 1.003; 1.008), history of stroke (OR 5.06; IC95% 1.04; 18.80) and arterial hypertension (OR 3.15; IC95% 1.38; 7.56). The antibiotic therapy delays not influenced the mortality or re-admission rate. In 24.3% the causative microorganism was identified and antibiotic treatment was inappropriate 19.6%. CONCLUSIONS: Hospital mortality rate was 5.8% and readmission rate was 5.7%. Variables associated with mortality differ from those associated with re-admission. The delay in the antibiotic initiation was not associated with a deleterious effect. Antibiotic therapy was inadequate in almost 20% of patients.