686. Impact of a Program of rational antibiotic use and new technology for early detection and treatment of gram negative bacteremia in a pediatric referral hospital
Session: Poster Abstract Session: They've Been Here a Billion Years! Pediatric Bacterial and Viral Infections
Thursday, October 27, 2016
Room: Poster Hall
Background: For the implementation of MALDI-TOF MS an intervention program for rational use of antibiotics is required Aim: To analyze the impact of an intervention program in addition to the MALDI-TOF in pediatric inpatients with gram-negative bacteremia

Methods:Before-after intervention study. Population: Children 1 m-18 years with gram-negative bacteremia. Poly-microbial bacteremia, burn and BMT patients were excluded. Main endpoint: Delay to adequate ATB treatment. Study period: Pre-intervention (Pre-I) 2014, post-I 2015. Intervention: Implementation of MALDI-TOF and VITEK2 data were reported in real-time by whatsapp to participants from microbiology, ID and pharmacy. Statistical analyses: STATA 9.0.

Results:

: 120 episodes were included (n= 60 Pre and PostI). Univariate analysis (PreI vs Post I) no statistical difference was observed: median age 33 months vs. 41 m , underlying disease 98 %vs. 95%,final diagnosis: Primary bacteremia 22% vs 12%, CVC-related bacteremia 35% vs 47%, secondary foci 43% vs 42%. Sepsis occurred in 53% vs 45%.Empirical treatment was adequate 97%vs 96%. Microbiological isolates: enterobacteria n=58 (14% BLES +) vs. 40 (15% BLES+), Pseudomonas aeruginosa 1 vs 10, other non-fermented 1 vs 6. Initial antibiotic treatment was modified 55% vs 62%. Median hours to adequate antibiotic treatment were 48 (IQR 36-48 hs.) vs 24 (IQR 24-48 hs), p=0.0016. Twenty-two vs 12 p. required PICU OR 0.42 (CI 95% 0.18-0.96) p=0.03. Median LOS in PICU were 25 (IQR 6-36) vs. 20 days (IQR 6-42) p=0.7, total median LOS were 29 (IQR 14-66) vs 16 days (IQR 13-35) p=0.05; 7 vs 5 patients died. Thirty-day mortality: 17% vs. 10% p=0.28. Median hours of inadequate vancomycin use :59 (IQR 37-85.5) vs. 27(IQR 22-47) p= 0.003, colistin 48(IQR 48-96) vs 24 (IQR19-24).

Overall PICU admission risk: univariate analysis: age <1 year OR 2.5 (1.07-6.3),p=0.038, median LOS days prior to bacteremia 15.5 vs. 2,p=0.002,median LOS total days 34 vs. 17, p=0.0002, mortality 30 days 10 vs 6 patients OR 5.5 (1.8-17),intervention program OR 0.42 (0.18-0.9), p=0.03. Multivariate analysis: LOS >14 days OR3.4 (1.4-8),p=0.007, Intervention program OR 0.4(0.17-0.9),p=0.04.

Conclusion: An intervention program with rapid detection of gram negative bacteremia was effective.

Silvina Ruvinsky, MD1, Moira Taicz, MD2, Soledad Martiren, MD2, Guadalupe Perez, MD2, Soledad Gonzalez, MD2, Vanesa Reijtman, Biochemical2, Alejandra Mastroianni, technician2, Romina Romero, MD2, Giannina Cinqui, Pharmacist2, Patricia Constanzo, Pharmacist2, Claudia Hernandez, Biochemical2 and Rosa Bologna, MD2, (1)Epidemiology and Infectious Disease, Hospital de Pediatría Dr, Juan P. Garrahan, Buenos Aires, Argentina, (2)Hospital de Pediatría Dr, Juan P. Garrahan, Buenos Aires, Argentina

Disclosures:

S. Ruvinsky, None

M. Taicz, None

S. Martiren, None

G. Perez, None

S. Gonzalez, None

V. Reijtman, None

A. Mastroianni, None

R. Romero, None

G. Cinqui, None

P. Constanzo, None

C. Hernandez, None

R. Bologna, None

Findings in the abstracts are embargoed until 12:01 a.m. CDT, Wednesday Oct. 26th with the exception of research findings presented at the IDWeek press conferences.