Multidrug-resistant organisms (MDRO) are very common and troublesome accompaniments of treating patients in intensive care units. Invasive treatment modalities (endotracheal tubes, intravascular and urinary catheters, among others), broad-spectrum antibiotic use and comorbidities are major contributing factors to colonisation and infection. Treatment of MDRO is often very difficult. Maasdorp et al. [1] have shown that the combination of colistin and tigecycline is considerably more effective in treating Acinetobacter baumannii (A. baumannii) infection than colistin alone in a retrospective cross-sectional case series. The number of patients enrolled in the study was small (16 in total), but there was considerably better ICU survival in the colistin-tigecycline treatment group (77.8%) than the colistin only group (28.6%), with survival to hospital discharge being 55.6% and 28.6%, respectively. A. baumannii is an organism with apparent variable pathogenicity, implying that treatment is not always necessary. The distinction between colonisation (with no need for antibiotic treatment) and invasive infection is very important. This study excluded patients with colonisation on the basis of absence of fever and other markers of active infection. Ntusi et al. [2] reported on a cohort of patients admitted to the intensive care units at Groote Schuur Hospital and identified 38 patients out 251 (15.1%) with positive cultures for A. baumannii as having colonisation. Bloodstream infection was present in over 70% of patients who died in this cohort and was an independent predictor of mortality. [2] Other risk factors for mortality included age, HIV-reactivity, low CD4 count, and multiple organ dysfunction occurring after onset of A. baumannii infection. Bloodstream infection was present in 44% of the Maasdorp cohort, but there was no apparent association with mortality, though this might reflect the small sample size. The challenge, and one that exists with many infections, is to identify patients with colonisation v. disease. That A. baumannii can be associated with severe pneumonia in a community-based setting, [3] its association with poor prognostic markers that are clinically meaningful, and a clear response to treatment in some individuals suggest that ignoring this organism, when identified, is an unsatisfactory approach (and which has been adopted by some clinicians). The identification of biomarkers that portend disease rather than colonisation, in large prospective studies, is an urgent unmet need in the field. Antibiotic treatment of infection is important, but is only one component of the complete management of MDRO infection. Measures to prevent the development and spread of MDRO are essential given the limited availability of effective antibiotics, particularly in an era when carbapenem-resistant enterobacteriaceae are becoming a greater problem than A. baumannii. Measures to prevent the spread of MDRO were discussed in a previous editorial. [4] Infection prevention and control measures that are essential include...