From: Selective decontamination and antibiotic resistance in ICUs
Author, year, country | Design (duration), setting | MDR species; population | Intervention – targeted/universal; study groups | Treatment efficacy | Resistance to SDD | Reported conclusion |
---|---|---|---|---|---|---|
Brun-Buisson and colleagues, 1989, France [22] | Prospective observational study (10 weeks) followed by RCT (8 weeks), medical ICU | MDR-E; ICU LOS >2 days and admission severity of illness score >2 (n = 210) | Universal: SDD = neomycin, polymyxin E and nalidic acid (no IV); (1) no SDD (observation, n = 124), (2) no SDD (RCT, n = 50), (3) SDD (RCT, n = 36) | ICU-acquired colonization with MDR-E: 19.6 % vs. 10 % vs. 2.9 % (P < 0.01 for 1 vs. 3); infection/colonization clinical site with MDR-E: 9 % vs. 6 % vs. 0 % | Rectal colonization with species resistant to SDD: 32 % vs. 58 % (P = 0.02, 2 vs. 3) | ‘… intestinal decontamination can be helpful to control outbreaks of multiresistant gram-negative bacilli, especially due to Klebsiella species, in the intensive care setting.’ |
Taylor and Oppenheim, 1991, USA [23] | Before–after study (2 + 2 months), multidisciplinary ICU | ESBL-KA; ICU LOS >48 hours (n = 33) | Universal: SDD = colistin, tobramycin and amphotericin (no IV); (1) no SDD (2 months, n = 15), (2) SDD (2 months, n = 18) | Colonization/infection ESBL-KA: 26.7 % vs. 0% | ‘No gram-negative aerobes resistant to the SDD drugs or ceftazidime emerged during the SDD regimen’ | ‘SDD appears to be a useful tool for eradicating outbreaks due to Gram-negative aerobic bacilli.’ |
Decré and colleagues, 1998, France [24] | Observational study (12 months), infectious disease ICU | ESBL-KP; (1) all patients admitted to ICU, (2) colonized and/or infected patients (n = 404) | Universal vs. targeted: SDD = erythromycin and polymyxin E (no IV); (1) universal SDD (7 months, n = 239), (2) targeted SDD (5 months, n = 165) | ICU-acquired colonization/infection with ESBL-K: 10.0 % vs. 9.1 %; ICU-acquired infection with ESBL-KP: 7.5 % vs. 3.6 % | Not reported | ’… prophylactic SDD failed to significantly reduce the incidence of acquisition of ESBL-producing strains.’ |
Agusti and colleagues, 2002, Spain [25] | Before–after study (2 + 2 months with 5 months between), ICU | Acinetobacter baumannii; patients with A. baumannii intestinal carriage, expected ICU LOS >5 days (n = 54) | Targeted: SDD = polymyxin E and tobramycin (no IV); (1) no SDD (n = 33), (2) SDD (n = 21) | A. baumanni colonization at ICU discharge: pharyngeal 78 % vs. 38 % (P = 0.03), rectal 91 % vs. 48 % (P < 0.001); A. baumanni in clinical samples: 81% vs. 45.5 % (P = 0.05) | ‘No resistance to colistin developed during the study’ | ‘… SDD may be beneficial, decreasing the intestinal reservoir in ICU patients with A. baumannii colonization […]. … if SDD is used in the setting of an A. baumannii outbreak, decontamination should not be restricted to the digestive tract, but applied also to the skin …’ |
Lubbert and colleagues, 2013, Germany [26] | Retrospective cohort (28 months), surgical ICU | KPC-2-KP; patients colonized/infected with KPC-2-KP (n = 90) | Targeted: SDD = colistin and gentamicin (no IV); (1) SDD (n = 14), six received concomitant IV antibiotic therapy for KPC-2-KP infection, (2) no SDD (n = 76), 22 received concomitant IV antibiotic therapy for KPC-2-KP infection | In-hospital mortality: 36 % vs. 45 %; decolonization (at day 21): 43 % vs. 17 % | Colistin: two patients receiving SDD (both also received colistin IV); gentamicin: five patients receiving SDD (three also received gentamicin IV) | “… SDD with gentamicin and colistin contributed to decolonization of KPC-2-KP in 6 of 14 cases (43 %) but revealed a substantial risk of rapid induction of secondary bacterial resistance to colistin and gentamicin.’ |