BackgroundOutpatient Parenteral Antibiotic Therapy (OPAT) is a safe, effective and beneficial practice but studies report 10–20% of readmissions rate. The risk factors for readmissions in OPAT have been investigated, although there are no clinical tools that allow us to predict these situations. The main goal of this study is to develop and validate a predictive model for readmission in OPAT patients.MethodsProspective study was conducted during 1 year (10/2012–09/2013), 1488 patients with OPAT were recruited at 8 units of Hospital at Home in Spain. Potential risk factors related to patient demographics, lead-time factors, clinical and microbiologic features were collected. We developed the prediction model in a derivation sample and after that, we validated this model in the validation sample. Sensitivity, specificity and area under the curve were obtained and the calibration capacity of the models were evaluated using the Hosmer-Lemeshow test (H-L).ResultsThe mean age of patients was 63 years (range 11–102), 58.74% men and most common diagnoses were urinary tract infections (23%). Our readmission rate during OPAT episode at home was 8.67% and the 30-days readmissions were 12.29%. The 72% of the readmissions during OPAT episode was related to the infectious pathology and 27.90% to the the patient’s comorbidity.The leading indicators for readmission were: gender, age, presence of caregivers, risk factor for infection, Charlsonand Barthel Index, microorganisme number, presence of multirresistent or micotic infection, venous access, antibiotic type and creatinine, proteine and leucocyte level at admissions. Finally, those factors included in the model were: antibiotic type (OR 3.93; IC 95% 1.90–8.11; P = 0.0002), presence of infection risk factor (OR 2.53; IC 95% 1.47–4.38; P = 0.001) and leucocytosis at admission (OR 2.21; IC 95% 1.32–3.71; P = 0.003). The AUC for the model was 0.72 (IC 95% 0.66–0.78) and the H-L value was 0.23. After the validation the AUC was 0.71 (IC 95% 0.64–0.78) and H-L value 0.9.ConclusionPatients at high risk of readmission during OPAT may be identified using predictive rules. This will allow us to implement measures that reduce the rate of readmissions and contribute to increase the safety of this therapy.Disclosures All authors: No reported disclosures.
Read full abstract