Validation of Drug Resistance in Pneumonia (DRIP) Score as Empirical Antibiotic Failure Predictor in Community-Acquired Pneumonia Patients in Cipto Mangunkusumo Hospital.

Background: The incidence of CAP due to Drug-Resistant Pathogen (DRP) requires broad-spectrum antibiotic therapy, Drugs Resistance in Pneumonia (DRIP) score can predict these cases. The use of the DRIP score can prevent antibiotic failure and long hospitalization, but validation is needed so that the DRIP score can be used according to the local community at Cipto Mangunkusumo National Central Public Hospital.
Methods: This research is a retrospective cohort study in CAP patients who were hospitalized during the period January 2019 to June 2020. Data were taken from medical records. Failure of empiric antibiotics occurs when one of these criteria is found: patient mortality, ICU transfer, and escalation of antibiotics as well as length of stay.
Results: 480 patients met the criteria. There were 331 patients (69%) with a DRIP score of <4 and 149 patients (31%) with a DRIP score of≥4. A total of 283 patients (59%) of antibiotic failures were detailed in 174 patients with a DRIP score <4 and 109 patients DRIP score ≥4. DRIP calibration using the Hosmer-Lemeshow test obtained p-value= 0.667 (p>0.05). AUC observations on the ROC curve obtained 0.651 (95% CI; 0.601-0.700).
Conclusion: The DRIP score has low accuracy performance and calibration value in predicting empirical antibiotic failure and poor discriminatory value.